THE HINDBRAIN NEURAL CREST AND THE DEVELOPMENT

OF THE ENTERIC NERVOUS SYSTEM

DE NEURAALLUST V AN DE ACHTERHERSENEN EN DE

ONTWIKKELING VAN DE DARMINNERVATIE

PROEFSCHRIFT

TER VERKRIJGING V AN DE GRAAD V AN DOCTOR

AAN DE ERASMUS UNIVERSlTElT ROTTERDAM

OP GEZAG V AN DE RECTOR MAGNIFICUS

PROF. DR. P.W.c. AKKERMANS M.LlT.

EN VOLGENS BESLUlT VAN HET COLLEGE VAN DEKANEN.

DE OPENBARE VERDEDIGING ZAL PLAATSVINDEN OP

WOENSDAG 2 FEBRUARI 1994 OM 13.45 UUR

DOOR

MARIA JOSEPHA HUBERTINA VANDER SANDEN

GEBOREN TE GELDROP

PROMOTIECOMMISSIE

Promotor:

Overige leden:

Co-Promotor:

Prof. Dr. J.e. Molenaar

Prof. Dr. F.G. Grosveld Prof. Dr. A.C. Gittenberger-de Groot Prof. Dr. med. B. Christ

Dr. J.H.C. Meijers

Dit proefschrift werd bewerkt binnen het Medisch Genetisch Centrum Zuid-West Nederland. binnen een samenwerkingsverband van het Instituut Kinderheelkunde en het Instituut Celbiologie en Genetica van de Faculteit der Geneeskunde en Gezondheidswetenschappen, Erasmus Universiteit Rotterdam.

Gedrukt bij Offsetdrukkerij Ridderprint B.V., te Ridderkerk

"We respond to fame's trumpeting of an individual scientist, but are often deaf to her

orchestrations; yet of all human activities the occupation of science is more like a symphony

than a solo ",

W. Grey Walter

Voor mijn ouders

Voor Wi!

CONTENTS

List of abbreviations

General Introduction and scope of this thesis

Chapter 1: Introduction to the neural crest

1.1. Evolutionary origin of the neural crest

1.2. The neural crest in amphibians

1.3. The neural crest in birds

1.4. The neural crest in mammals

1.5. Patterning of the rhombencephalic neural crest

1.6. In vitro studies of the neural crest

1.7. Clinical disorders of the neural crest

1.8. Conclusions

1.9. References

Chapter 2: Introduction to the enteric nervous system

2.1. Evolutionary aspects of intestinal motility

2.2. Structure and 'ultrastructure of the enteric nervous system

2.3. Development of the enteric nervous system in amphibians

2.4. Development of the enteric nervous system in birds

2.5. Development of the enteric nervous system in mammals

2.6. Clinical disorders of the enteric nervous system

2.7. Conclusions

2.8. References

Chapter 3: The experimental work

3.1. Introduction to the experimental work

3.2. Ablation of various regions within the avian vagal neural crest

has differential effects on ganglion formation in the fore-, mid­

and hindgut.

Peters~van der Sanden, M.J.H., Kirby, M.L., Gittenberger-de

Groot, A.C., Tibboel, D., Mulder, M.P., and Meijers, c.; Dev.

Dyn. 196:183-194, 1993.

8

9

15

16

18

22 24

27

29 31

32

43

46 50

50

52

54 57

57

65

69

3.3. Differential effect of retinoic acid on ectomesenchymal and 81 ganglionic derivatives of the hindbrain neural crest in chicken

embryos.

Peters-van der Sanden, M.J.H., Vaessen, M.J., and Meijers, c.; submitted.

3.4. Colonization characteristics of enteric neural crest cells: 97 embryological aspects of Hirschsprung's disease.

Meijers, c., Peters-van der Sanden, M.J.H., Tibboel, D., van

der Kamp. A. W.M .• Luider. T.M .. and Molenaar. J.e.; J. Ped.

Surg. 27:811-814. 1992.

3.5. Regional differences between various axial segments of the 101 avian neural crest regarding the formation of enteric ganglia.

Peters-van der Sanden, M.J.H., Luider, T.M., van der Kamp,

A. W.M.. Tibboel. D .• and Meijers. e.; Differentiation 53: 17-24.

1993.

3.6. Characterization of HNK -1 antigens during the formation of the 109 avian enteric nervous system.

Luider, T.M., Peters-van der Sanden, M.J.H., Molenaar, J. e., Tibboel, D., van der Kamp, A. W.M., and Meijers, e.; Development 115:561-572. 1992.

3.7. Patterns of malformations and dysmorphisms associated with 121 Hirschsprung's disease: an evaluation of 214 patients.

van Dommelen, M. W., Peters-van der Sanden, M.J.H.,

Molenaar, J.e., and Meijers, C.; submitted.

3.8. General discussion 139 I. Segmentation within the neural crest regarding enteric

nervous system formation 139 2. Migration of neural crest cells to the gut 141 3. Migration of neural crest cells through the gut 144 4. Homing of neural crest cells in the gut 144 5. Differentiation of neural crest cells and the role of the

enteric microenvironment 145 6. Clinical implications 147

3.9. References 148

Summary 151

Samenvatting 153

Curriculum Vitae 155

List of publications 157

Dankwoord 159

List of abbreviations

CNS:

Dil:

E .. :

ENS:

HNK-l:

HSCR:

Is:

RA:

S:

SI:

Central Nervous System

1,I-dioctadecyI-3,3,3' ,3' -tetramethylincarbo-cyanine perchlorate

Embryonic Day

Enteric Nervous System Human Natural Killer cell (monoclonal antibody)

Hirschsprung's disease

Lethal spotted

Retinoic Acid

Somite

Piebald Lethal

GENERAL INTRODUCTION AND SCOPE OF TInS THESIS

The wonder of things is the beginning of knowledge, as was already stated by Aristotle, the

fIrst embryologist known to history. Embryology has remained a source of wonder ever since. It all starts with the fusion of the female egg and the male sperm. Sperm cells were first

described by Antonie van Leeuwenhoek (1632-1723) in 1678, who believed them to be

parasitic animals present in the male semen, that had nothing to do with reproduction. Nicolas Hartsoeker (1656-1725), the other discoverer of sperm believed that the entire embryonic

individual lay preformed within the head of the sperm, as depicted in his famous homunculus

(Fig. 1). The fIrst evidence for the existence of the female egg was presented by Reinier de

Graaf (1641-1673), although the egg itself was only described in 1827 by Karl von Raer

(1792-1876). The actual fertilization process was observed only a century ago by Herman Fol,

a Swiss zoologist.

The development of one single cell (the zygote) into a complex organism consisting

of many different cell-types and organs, entails a number of processes including cell migration, proliferation, differentiation, growth and pattern formation. During the cleavage

stage, the zygote divides a number of times leading to the formation of a blastula. This

blastula is then transformed, due to a series of complex

cell movements, into a gastrula, a three-layered embryo

consisting of endo-, meso- ,and ectoderm, in which the

blue print for segmentation is laid out. Also at this

stage, the primitive gut is formed. During neurulation,

the mesoderm induces part of the overlying ectoderm to

form a neural plate, which subsequently develops into

the neural tube. During this process, the vertebrate

neural crest arises on the dorsal aspect of the neural

tube (Fig. 2). The neural crest is a migratory population

of cells, which was first described in chicken embryos

by Wilhelm His (1831-1904) in 1868. He described it

as a band of particular material, which he called the

Zwischenstrang, lying betwe,en the presumptive

epidermis CHomblatt') and the neural tube. Neural crest

cells emigrate from the neural tube, either prior to

fusion of the neural folds (amphibia and mammals), or

shortly following its closure (birds). After initiation of

migration, neural crest cells embark upon migratory

pathways, characteristic of their axial level of origin,

Figure 1: The human infant preformed in the sperm: Nicolas Hartsoeker's homunculus. (From N. Hartsoeker, 1694, Essai de Dioprique.)

9

and give rise to a large variety of derivatives. These include ganglionic derivatives, such as

the neurons and supportive cells of the peripheral nervous system, ectomesenchymal

derivatives. such as the bony and cartilaginous structures of the head, and adrenomedullary

and pigment cells (Table 1). Although a great deal is already known about the migratory

pathways and range of potential derivatives, there are still many open questions concerning

neural crest cell migration and differentiation.

The hindbrain or rhombencephalic neural crest gives rise to the neurons and supportive

cells of the enteric nervous system (ENS) (Yntema and Hammond, 1954; LeDouarin. 1982).

This local nervouS apparatus is embedded in the wall of the gut and is responsible for

peristaltic contractions. These are true coordinated reflexes consisting of proximal contractions

and distal dilatations, resulting in a craniocaudally directed movement of the contents of the

gut. The ENS is considered to be a

separate division of the autonomic

nervous system (Langley. 1921). and

differs from the other components of

the peripheral nervous system. i.e. the

sympathetic and the parasympathetic

ganglia. in that it can mediate reflex

activity independently of the central

nervous system (eNS) both in vivo

(Bayliss and Starling. 1899: Bayliss

and Starling. 1900) and in vitro

(Trendelenburg. 1917). This indicates

that the ENS is a self-contained

nervous system, the only such system

in the periphery.

Although the gut can function

independently from the CNS. it does

not normally do so (Roman and

Gonella. 1981). Both sympathetic and

parasympathetic nerves innervate the

bowel (Kuntz. 1963). In addition.

sensory neurons located in the vagal

and dorsal root ganglia project to the

gut (Ewart. 1985). Thus. peristaltic

activity can be influenced by the CNS

in the normal control of gastro­

intestinal function. Apart from its role

10

Neural crest

" • Neural plate

/'

Neural folds

~~c=-c=> 8 c=>

Neural crest

/' ~

@ffi(s -.::: > e c~

Figure 2: Schematic diagram illustrating transverse sections through embryos at progressive stages of neural tube closure. The neural folds form from the neural plate under the influence of the notochord (No). The neural folds come together and close to form the cylindrical neural tube (NT). Neural crest cells emerge from the dorsal neural tube and migrate extensively. Som=somite. (Adapted from Selleck et at.. 1993)

Table 1: Derivatives of the different regions of the neural crest

Neural crest region Mesencephnlic Anterior Rhombencephalic Posterior Rhombenccphalic Truncal

Pharyngeal arches I 1 and 2 3, 4 and 61 from MO-S5' Caudal to somite 5 somites

Ectomesenchymai derivatives

Skeletal Nasal. orbitary skeleton Reichert's cartilage Hyoid cartilage Trabeculae Parietal bones Laryngeal cartilages Sphenoid capsule Palatine, Otic capsu Ie Cranial vault .Maxilla Anterior squamosal bone Skeleton of the lower jaw (including Frontal bone Meckel's cartilage) Rostrum of the pamsphcnoid Pterygoid

connective I l\'leninges of the forebrain Tooth buds Mesenchymal components of muscle Muscles and connective tissue Muscles and connective tissues of aortic arch arteries, aorto~

of the face the head and neck region pulmonary septum, semilunar valves, thymus, thyroid, and parathyroids

Ganglionic derivativcs Cranial nerves I~IV Trigeminal (V) root ganglion Glossopharyngeal (IX) root Dorsal root ganglia (neurons and glial Facial (VII) root ganglion ganglion Sympathetic ganglia cells) Vagal (X) root ganglion

Cardiac ganglia Enteric ganglia

Other derivatives Pigment cells Pigment cells Pigment cells Pigment cells Carotid body type I cen~ Adrenal medulla Calcitonin-producing cells

• from the level of the mid-otic vesicle down to the caudal boundary of somite 5

in intestinal motility, the ENS is also involved in controlling gastrointestinal immunity. the

resorption and secretion process, and intestinal blood-flow (Gershon, 1981; Cooke, 1986;

Felten et al., 1988).

Scope of this thesis

Knowledge on the cellular and molecular mechanisms of the development of the hindbrain

and its neural crest is rapidly increasing. but still far from complete. The neural crest of the

posterior rhombencephalon (rhombomeres 6-8) contributes to a number of derivatives, partly

ectomesenchymal, such as cells in the outflow tract of the heart (in the media of the aortic

arch arteries, the aorto-pulmonary septum and the semilunar valves) and the mesenchymal

components of the thymus and parathyroids, partly ganglionic, such as the nervous system of

the digestive tract and the cardiac ganglia. The aim of the experimental work described in this

thesis is to extend our understanding of the cellular and molecular mechanisms of the

development of the ENS. In order to identify and characterize the neural crest cells involved.

segments of hindbrain neural crest were ablated in chicken embryos and the effects on ENS

development studied. A specific segment adjacent to somites 3-5 was further characterized

in in vitro cultures and in an in vivo colonization assay. To identify homing andlor

differentiation signals for neural crest cells, the enteric microenvironment of aneura! and

neural gut was compared. A brief overview of the neural crest and the ENS in various species

is given. Finally, the current data on the cellular and molecular mechanisms of ENS

development are discussed.

12

Chapter 1

Introduction to the neural crest

1.1. The evolutionary origin of the neural crest

The evolutionary appearance of the neural crest and of the epidermal neurogenic placodes

(local thickenings of the ectoderm. which also contribute to the vertebrate nervous system)

has been closely tied to the development of the brain. skull. sensory apparatus and muscular

pharynx of the vertebrate head (Gans and Northcutt. 1983). Maisey (1986) stated that the

neural crest is the quintessential characteristic of vertebrates: the craniate features resulting

from the presence of the neural crest. both directly, through the multitude of tissues produced

by the neural crest and indirectly. through the role of the neural crest as an inducer of tissues

arising from other germ layers.

The transition from prevertebrates to vertebrates is considered not to be a

macroevolutionary change, but rather to represent a series of gradual changes (Gans and

Northcutt. 1983; Northcutt and Gans. 1983). The deuterostome invertebrate was a small, filter­

feeding, marine organism, who used multiple ciliary bands to drive propUlsion. filtration and

food transport. It had no specialized sense organs or brain, but a distributed epidermal nerve

plexus. coordinating motor response on the basis of local sensory input. Transition to

protochordates entailed the development of the notochord, flanked by muscles capable of

undulating the notochord, resulting in more effective locomotion. This red:uced the need for

a ciliated surface, which led to redundancy of part of the sensory capacity and thus of part

of the epidermal nerve plexus. This partly redundant epidermal nerve plexus might have

evolved into the neural crest in vertebrates.

The shift from passive drift toward powered and oriented movement presumably

established the selective advantage for an improvement of gas exchange, which until then had

depended solely on diffusion. In agnatha. the first Gawless) vertebrates. this led to the

development of a branchial arch system. through which water containing both oxygen and

food particles was pumped. This branchial arch system was supported by cartilage, which was

the first discrete neural crest tissue in ontogeny that arose from transformation of the

epidermal nerve plexus (Maisey. 1986; Hall, 1992).

Modification of the first pair of branchial arches into jaws in gnathostomes, was an

important step in the transition from filter-feeders to predators. Prey detection improved with

the development of electroreceptors, associated with dentine and enamel. the first hard tissues

that derived from the neural crest. In order to further stabilize these new signal sources.

secondary specializations developed from the neural crest in the form of bone (and perhaps

also cartilage). with the sense organs resting in or among bony plates (capsules). Thus. the

neural crest, which was initially a sensory tissue. now also became involved in skeletogenesis.

After the development of lungs, which accompanied a terrestrial life style. the branchial arch

system evolved into the more complicated pharyngeal arch system found in higher vertebrates.

The first evidence presaging the neural crest as the origin of vertebrate pigment cells

15

Fossil record

Protochordate Hypothetical

protovert~ebrate A9naytha ~t~n{:::mes Vertebrae Postotlc skull

Etectroreception External bone

Predation Paired external sense organs

Start of new head /' Brain

Pharyngeal breathing Sranchlomerlc muscle Gill capillaries

I Start of ENSI Cartilage

" Active dispersing Segmented muscle Y Notochord

Filter feeding Collagenous ciliated pharynx: /' Integumentary exchanger

Epidermal nerve plexus

Figure 3: Hypothesized structural and functional transitions in vertebrate evolution. The postulated functional states (capitals) precede the modified structures (lower case letters) involved with them. (Adapted from Gans and Northcutt. 1983)

comes from Tunicates (subphylum Urochordata), who posses a line of pigment cells as a band

along the dorsal aspect of the neural tube. These cells are the first and only cells in Tunicates,

which are not fully determined and can switch to the muscle cell phenotype, giving further

indications that these cells represent derivatives of protoneural crest cells. This shows that all

neural crest-derived tissues, characteristic for vertebrates, developed more or less

independently in various protochordate and early vertebrate species (Fig. 3).

1.2. The neural crest in amphibians

At the closure of the neural folds, amphibian neural crest material, lying in the ridges on each

side, fuses together in the midline and constitutes a wedge-shaped cell-mass, which separates

the two halves of the neural tube dorsally. Neural folds arise at the boundary between

epidermis and neural plate, which is created after neural induction through the notochord

(Jacobson. 1981). Neural folds. however. also arise at experimentally created boundaries

between epidermis and neural plate and these new neural folds also give rise to neural crest

cells (Moury and Jacobson, 1989), indicating that the notochord, although responsible for

neural plate induction, is not directly involved in the induction of the neural crest. Neural

crest cells originate from both the neural plate and the epidermis in both naturally occurring

as weD as in experimentally induced neural folds (Moury and Jacobson. 1990). Soon after

formation, neural crest cells leave the neural tube, migrate throughout the embryo, and give

16

rise to a large number of derivatives.

Cranial neural crest

Migration of neural crest cells in the head starts in the mesencephalon before closure of the

neural tube. In the prosencephalon, rhombencephalon and trunk, neural crest cells leave the

neuroepithelium only after separation of the neural tube from the overlying epidermis (Raven,

1931). Many studies have been performed to create a fate-map of the amphibian cranial neural

crest, using either neural crest ablation (Stone, 1922; Stone, 1926; Stone. 1929; Horstadius

and Sellman. 1946). intra- and inter-specific neural crest chimeras (Horstadius and Sellman,

1946; Chibon, 1964; Chibon, 1966), or vital dye labelling of neural crest cells (Horstadius and

Sellman, 1946; Collazo et al., 1993). These studies demonstrated that a limited portion of the

cranial neural crest, from the level of the caudal prosencephalon to the posterior

rhombencephalon, contributes to the cranial and visceral skeleton. The most rostral neural

crest cells do not contribute to the skeleton, although grafting experiments show that they

have the potential to chondrify upon contact with endoderm (Seufert and Hall, 1990).

Horstadius and Sellman (1946) showed that cranial neural crest cells grafted into the trunk

region, failed to chondrify unless pharyngeal endoderm was included in the graft, further

illustrating the need for tissue interactions in neural crest cell differentiation into cartilage.

Trunk neural crest

For the axolotl embryo, it is generally agreed upon that trunk neural crest cells migrate along

three major pathways: dorsally, where they form the mesenchyme of the dorsal fin, laterally

(between somites and epidermis), where they give rise to pigment cells, and ventrally

(between somites and neural tube), where they form the elements of the peripheral nervous

system (Schroeder, 1970; MacMillan, 1976; Vogel and Model, 1977; Lofherg et al., 1980).

In Xenopus laevis embryos, however, different opinions exist on neural crest cell migration,

especially on the extend of the lateral pathway. Whereas Krotoski et al. (1988) and Collazo

et al. (1993) described a minor lateral pathway, Epperlein et al. (1988) could find no evidence

for cells migrating laterally. Neural crest cells migrating along the ventral route in Xenopus

embryos, were present between the neural tube and the posterior half of the somites, thus

showing a metameric pattern, whereas few cells were present within the somites (Krotoski et

aI., 1988). For the caudal trunk region, two additional pathways into the ventral fin have been

described (Collazo et aI., 1993). One group of cells migrates circumferentially within the fin.

while another progresses ventrolaterally to the anus before populating the ventral fin. This

latter group of cells passes through the enteric region. where they can be found temporarily.

The onset of migration in the trunk is temporally and regionally well coordinated,

resulting in a wave of initiated neural crest cell migration passing along the body axis in a

craniocaudal direction (Detwiler, 1937; Bancroft and Bellalrs, 1976; Tosney, 1978; Lofherg

17

et al., 1980). Little, however, is known about the mechanisms regulating this coordinated

onset of cell migration, although there is evidence indicating that factors intrinsic to the neural

crest cells (Newgreen and Gibbins, 1982) as well as factors in the embryonic environment

(L6fberg and Nynas-McCoy, 1981; Erickson and Weston, 1983) may playa role. Both

fibronectin and tenascin are present in the extracellular matrix lining the neural crest

migratory pathways at the time when neural crest cells are actively migrating (Epperlein et

al., 1988). Bur, whereas fibronectin is already present before the onset of neural crest

migration, the appearance of tenascin is correlated with the initiation of migration, suggesting

that an interaction between these two extracellular matrix components could be important in

regulating the onset and pathways of neural crest cell migration. In the axolotl embryo, it has

been shown that the subepidermal extracellular matrix, which forms a substrate for cell

locomotion, initiates and regulates the onset of neural crest cell migration along the lateral

pathway (L6fberg et al., 1985). In the white axolotl mutant, trunk pigmentation is restricted.

because the epidermis is unable to support subepidermal migration of pigment cells, due to

a retarded maturation of the extracellular matrix (L6fberg et al., 1989).

1.3. The neural crest in birds

The avian neural crest forms during the second day of embryonic development (E2, stage 8-

16, Hamburger and Hamilton, HH; 1951) along the entire antero-posterior axis of the embryo.

Cranial Trunk

~ -- ............ ,----""-k///'//h'///';'/,////'/~~//.i////.'////////////!I > Neural crest

o ITl@J@]~@l~[fJ~~GQ Somites Oto

Cardiac Neural crest

Figure 4: Diagram illustrating the various parts of the neural crest. The cranial neural crest extends from the mid-diencephalon to somite 5 and trunk neural crest begins at somite 5 and extends to the caudalmost limit of the neural tube. The cardiac neural crest, from the level of the mid-otic vesicle down to the caudal boundary of somite 3, forms a transitional region.

18

Depending on the axial level of origin, the neural crest can be divided into several different

regions (Fig. 4), each giving rise to specific derivatives. Numerous studies have been

conducted to trace neural crest cell migration and to construct a fate-map of the neural crest

(see LeDouarin, 1982 for review). In these studies, a number of different techniques were

used: many studies used surgically created chimeric embryos, transplanting either 3H_

Thymidine labelled (Weston, 1963; Johnston, 1966) or quail-derived neural crest cells

(LeDouarin, 1982; Noden, 1988) to chicken embryos; other studies used the monoclonal

antibody HNK-1 (or NC-1) to study early neural crest cell migration (Vincent et aI., 1983;

Vincent and Thiery, 1984); more recent studies used microinjection of vital dyes, like 1,1-

dioctadecyl-3,3,3' ,3' -tetrarnethylincarbo-cyanine perchlorate (Dil) (Lumsden et al., 1991;

Serbedzija et al., 1991) and lysinated rhodamine dextran (Bronner-Fraser and Fraser, 1988),

or retroviral markers (Stocker et al., 1993), to trace neural crest cell migration. Major

advantages of these latter techniques are that they do not require surgical intervention and

allow for lineage analysis by injecting single neural crest cells.

Cranial neural crest

The cranial neural crest entails the neural crest overlying the diencephalon, mesencephalon

and rhombencephalon, and ends at the caudal boundary of the fifth somite. The

prosencephalon of avian embryos does not give rise to neural crest cells in contrast to the

amphibian and mammalian forebrain. The mesencephalic neural crest cells are the first to start

migrating at stage 8, after closure of the neural tube. Migration proceeds in both anterior and

posterior direction. Neural crest cells of the di- and mesencephalon initially remain dorsal to

the neural tube and shift rostrally over the roof of the forebrain. Subsequently, they disperse

rostrally and laterally between the prosencephalic neuroepithelium and the surface ectoderm

to fonn all of the frontonasal mass and much of the periocular mesenchyme. giving rise to

cranial ganglia and contributing to the facial skeleton.

Mesencephalic neural crest cells start migration in a region rich in fibronectin, laminin,

heparan sulphate proteoglycan and tenascin (Krotoski et al., 1986; Bronner-Fraser and Lallier,

1988). together fOlming a dense matrix surrounding and interdigitating with premigratory

neural crest cells. In vivo perturbation experiments, in which antibodies against these

extracellular matrix molecules or their receptors were injected in the mesencephalic region

of early chicken embryos, showed that these extracellular matrix molecules were all involved

in cell migration (Matthew and Patterson, 1983; Bronner-Fraser, 1985; Bronner-Fraser, 1986;

Poole and Thiery, 1986; Bronner-Fraser and Lallier, 1988). Microinjection of the monoclonal

antibody HNK-I was also found to disturb cranial neural crest cell migration (Bronner-Fraser,

1987). This monoclonal antibody recognizes a carbohydrate epitope, which was first described

on human natural killer cells. and was found to react with early migrating neural crest cells

19

and their neuronal derivatives (Vincent et al., 1983). Since then, however, it was found to be

not entirely specific for these cells. It was found to be present on a large number of molecules

all involved in cell-cell or cell-substrate adhesion (Kruse et a1.. 1984). Injection of antibodies

against the cell adhesion molecules N-CAM and N-Cadherin also perturbs early cranial neural

crest cell migration (Bronner-Fraser et al., 1992), further indicating that the balance between

cell-cell and cell~matrix adhesion may be critical for cranial neural crest cell migration.

The rhombencephalic, or hindbrain, part of the vertebrate central nervous system is

segmented, consisting of eight consecutive rhombomeres (Vaage, 1969; Keynes and Lumsden,

1990)., Rhombencephalic neural crest cells migrate predominantly along a dorsolateral

pathway underneath the surface ectoderm and populate the pharyngeal arches, in a way that

neural crest cells of each rhombomere populate a particular pharyngeal arch which

subsequently gives rise to specific derivatives. Whether or not rhombomeres 3 and 5 give rise

to neural crest cells is not entirely clear (Lumsden et al .• 1991; Graham et al .• 1993; Sechrist

et a1.. 1993). In this region increased cell death has been described (Lumsden et al .. 1991;

Jeffs et a1.. 1992; Jeffs and Osmond. 1992). Table 1 shows the relationship between neural

crest cells of each rhombomere, the arch they populate, and the derivatives they will give rise

to. Neural crest cells of the anterior rhombencephalon, ranging from the mid-hindbrain border

to the mid-otic vesicle, comprising rhombomeres 1-5, populate the first two pharyngeal

arches, which mainly contribute to the craniofacial skeleton, the hyoid and to cranial ganglia.

The posterior rhombencephalic neural crest, overlying rhombomeres 6-8, has its anterior

boundary at the level of the ntid-otic vesicle. The posterior boundary of rhombomere 8 is less

clear, but is thought to lie between the fifth and sixth somite. This region of the neural crest

populates pharyngeal arches III. IV and VI, and gives rise to both ectomesenchymal and

ganglionic derivatives. Ectomesenchymal derivatives mainly entail cells in the outflow tract

of the heart and of the carotid and ultimobranchial body. and the mesenchymal component

of the thymus and parathyroids. Ablation experiments showed that these ectomesenchymal

derivatives originate mainly from a subregion of the posterior rhombencephalon, from the

level of the ntid-otic vesicle down to the caudal boundary of sontite 3 (Kirby et al .. 1983;

Bockman and Kirby, 1984). This so-called cardiac neural crest also gives rise to ganglionic

derivatives, such as cranial and cardiac ganglia (Kirby and Stewart, 1983). Migration of the

cardiac neural crest has been studied extensively, both in whole-mounts (Kuratani and Kirby,

1991) and in quail-chick chimeras (Miyagawa-Tontita et al .• 1991). These studies showed that

cardiac neural crest cells migrate predominantly along a dorsolateral pathway, temporarily

arresting to form the circumpharyngeal crest before populating the third, fourth and sixth

pharyngeal arches. The ntigration pathways of the rhombencephalic neural crest cells caudal

to the cardiac crest (adjacent to somites 4-7) have not been studied in detail.

The enteric nervous system, a ganglionic derivative of the posterior rhombencephalic

neural crest, was found to derive from the vagal neural crest (LeDouarin and Teillet. 1973).

20

This vagal neural crest partially overlaps the posterior rhombencephalic and cardiac crest, but

extends somewhat more caudally. down to the caudal boundary of somite 7. In this thesis, we

present evidence that a specific subregion within the vagal neural crest, adjacent to somites

3-5, is primarily responsible for ENS development (Peters-van der Sanden et al., 1993).

Trunk neural crest

Cells from the trunk crest region, ranging from the anterior boundary of somite 6 down to the

level of the tail bud, migrate along two major pathways (Serbedzija et al .. 1989; Bronner­

Fraser et aI .. 1991). The majority of the cells migrate along a ventrolateral pathway, through

the somites, and give rise to the neurons and supportive cells of dorsal root and sympathetic

chain ganglia, and, at the' level of somites 18 to 24. to neuroendocrine cells of the adrenal

medulla (LeDouarin and Teillet, 1973). Other trunk neural crest cells migrate along a

dorsolateral pathway underneath the ectodenn and give rise to pigment cells.

Neural crest cells migrating along the ventrolateral pathway, emigrate from the neural

tube in a continuous antero-posterior stream, but subsequently become restricted to the rostral

half of each somite (Rickmann et al., 1985; Teillet et al., 1987). This metameric migration

pattern results in the segmental arrangement of the dorsal root and sympathetic chain ganglia,

which form aligned with the rostral half of each trunk somite (Teillet et al .. 1987; Lallier and

Bronner-Fraser, 1988). The metameric pattern of migration was shown to be inherent to the

somites. After a 1800 antero-posterior rotation of the segmental plate, which will give rise to

the somites, neural crest cell.s migrated through the portion of the somite that was originally

rostral, but was now caudal with respect to the orientation of the embryo (Stem et al., 1991b).

Replacement of normal somites with only rostral halves leads to the fonnation of large,

unsegmented ganglia (Kalcheim and Teillet, 1989; Goldstein and Kalcheim, 1991). Motor and

sensory axons that grow from the neural tube also specifically go through the rostral half of

the somites (Keynes and Stern, 1984). The preferential migration through the rostral half of

the somites of both neural crest cells and axons could be caused either by stimulatory factors

present in the rostral half or by inhibitory factors present in the caudal half of each somite.

Most studies perfonned until now point to the presence of inhibitory factors in the caudal

half. although th~ definite candidate has yet to be identified (Stern and Keynes, 1987; Norris

et al .. 1989; Stern et al .. 1989; Ranscht and Bronner-Fraser, 1991). The first five somites

adjacent to the posterior rhombencephalon are completely inhibitory for ganglion formation,

although neural crest cell migration into these somites has been observed (Lim et al., 1987;

Teillet et al., 1987). This indicates that the somitic mesenchyme can not only inhibit neural

crest cell migration. but also regulates gangliogenesis.

The notochord and the perinotochordal matrix also inhibit neural crest cell migration

(Newgreen et aI., 1986; Pettway et al., 1990; Stern et al., 1991a). Fixation of the notochord

or trypsin treatment abolishes this inhibitory effect, indicating that the responsible substance

21

is proteinaceous (pettway et al., 1990). Although the notochord is known to induce ventral

neural tube structures (van Straaten et al., 1985; Jessel et al., 1989: van Straaten et al., 1989),

it does not prevent formation of the dorsal neural crest (A.rtinger and Bronner-Fraser, 1992a).

Additional evidence that neural crest cell migration takes place through all available spaces

unless it is restricted by inhibitory cues, comes from a study in which the neural tube was

rotated around its dorso-ventral axis (Stern et al .. 1991a). The neural crest cells now started

migrating dorsally, showing that they do not possess intrinsic directionality in their migration,

but rather exploit all those areas accessible to them and that do not inhibit their migration.

Emigration of trunk neural crest cells from the neural tube at a certain axial level

occurs during a prolonged period of time. The first cells that leave the neural tube embark

on the ventrolateral pathway, while the latest emigrating cells migrate dorsolaterally. Even

within the ventrolateral pathway, there is a difference between the developmental potential

of early and late emigrating neural crest cells (Weston and Butler. 1966). Whereas early

emigrating cells migrate far ventrally to give rise to sympathetic ganglia. later emigrating cells

remain in more dorsal positions and give rise to dorsal root ganglia. Transplantation of • old'

crest into a 'young' host and visa versa has shown that the last cells to leave the crest Cold'

crest) have the same range of developmental capabilities in a 'young' host as early migrating

cells. Within the host environment, however. some temporal alterations occur which

progressively limit the distal migration of neural crest cells. Trunk neural crest cells that

migrate along the dorsolateral pathway show a one-day delayed emigration from the neural

tube compared to ventrolaterally migrating cells at the sarne axial level (Erickson et al .. 1992).

This delayed migration is controlled by the dermatome, removal of which results in a quicker

embarkment onto the dorsolateral pathway. Once neural crest cells have entered the

dorsolateral pathway, they colonize it very rapidly. These late-emigrating neural crest cells

are partially restricted in their developmental potential, being no longer able to differentiate

into adrenergic cells neither in vivo nor in vitro (Artinger and Bronner-Fraser. 1992b). These

results show that the contribution of trunk neural crest cells to their derivatives occurs in a

ventral to dorsal progression. with the precursors for pigment cells being the last to exit the

neural tube.

1.4. The neural crest in mammals

Study of the neural crest in mammalian embryos has traditionally been based on descriptive

morphology and extrapolation from other vertebrates. because mammalian embryos are not

amenable to transplantation techniques. Although this approach may be valid for the trunk.

it has serious limitations for the cranial region. because mammalian embryos have taken

cranial specialization a stage further than other vertebrates. This becomes increasingly clear

22

as development proceeds. with the enormous expansion of the telencephalon to form the

cerebral hemispheres with a true neocortex. Although there is considerable variation in the

pattern of migration among mammalian embryos as a group. the basic principles of

craniofacial development are very similar.

Cranial neural crest

Although some aspects of cranial neural crest cell migration are similar to those observed in

avian embryos, there are variations in the exact pathways and timing of migration, and in the

axial levels that contribute to the neural crest. Cranial neural crest cells in mammals emigrate

from the neural folds prior to neural tube closure (Verwoerd and van Oostrom, 1979; Nichols,

1981; Nichols, 1986), Closure of the neural tube starts at the 7-somite stage at the level of

somites 4 and 5, and progresses both rostrally and caudally. In the cranial region closure is

complete at the 16-somite stage.

Recently, study of neural crest cell migration in the mammalian head has been made

feasible through injections ofWGA-Au (wheat-germ agglutinin-gold) or DiI into the amniotic

cavity (Smits-van Prooije et ai" 1988; Serbedzija et aI" 1992), Migration starts at the 5-somite

stage in the rostral hindbrain (rhombomere 1). followed by migration in the midbrain and

caudal hindbrain and finally in the forebrain, In the mouse embryo, neural crest cells of the

forebrain migrate ventrally in a contiguous stream through the mesenchyme between the eye

and the diencephalon. Midbrain neural crest cells migrate through the mesenchyme as

dispersed cells, which differs from the subectodermal stream observed in the midbrain of

avian embryos (LeDouarin, 1982), In the hindbrain of mouse embryos, neural crest cells

migrate in three subectodermal streams, each extending into the distal portions of the adjacent

pharyngeal arches similar to the migration described in avian embryos (Lumsden et al., 1991).

In the rat embryo, the forebrain does not give rise to neural crest cells, similar to avian

embryos (Noden, 1975; Tan and Morris-Kay, 1985; Smits-van Prooije et aI., 1988). In the

midbrain of the rat embryo neural crest cells migrate as dispersed cells similar to mouse

embryos. In the rat hindbrain, crest cells migrate ventrally through the mesenchyme as

dispersed cells (Tan and Morris-Kay, 1986). which differs from the subectodermal streams

observed in mice and avians. In the mouse embryo, all 8 rhombomeres give rise to neural

crest cells (Serbedzija et aI., 1992), whereas in chicken embryos there seem to be no neural

crest cells migrating from rhombomeres 3 and 5 (Lumsden et al., 1991). Recent evidence,

however. suggests that rhombomere 3 and to an even greater extent rhombomere 5 generate

neural crest cells in avian embryos as well (Sechrist et aI" 1993).

Mammalian neural crest cells were found to be able to participate in normal embryonic

development after microinjection into post-implantation embryos (Jaenisch, 1985). Neural

crest cells and fibroblasts injected into the amniotic cavity of early mouse embryos become

dispersed into the cranial mesenchyme along normal migratory pathways. in contrast' to

23

hepatoma cells and latex beads (Chan and Lee, 1992), This indicates that the ability to

disperse along neural crest migratory pathways correlates with the ability to respond

adequately to signals in regions of active cell migration and rules out passive displacement.

Injection at different time-points in development showed that the migratory environment

changed concomitantly with an increase in developmental age, limiting the extend of

migration. Such a change in migratory environment related to developmental age was also

described for the trunk neural crest of avian embryos (Bronner-Fraser and Cohen, 1980).

Trunk neural crest

Mammalian trunk neural crest cells migrate along two major pathways. one ventrolateral

through the anterior part of the somites, the other dorsolateral underneath the ectoderm.

Migration is very similar to avian trunk neural crest cells except for the timing. Neural crest

cells appear on the dorsal surface of the neural tube 2-4 somites rostral to the most recently

formed somite, whereas in avian embryos neural crest cells appear about 5 somites rostral

(Erickson et aI" 1989; Serbedzija et aI" 1990), The ventral pathway is segregated into two

phases of migration (Serbedzija et al" 1990), The early phase starts before E9 and ends at

E9.5 and consists of a stream of cells within the rostral sclerotome, which extends ventrally

to the region of the sympathetic ganglia. The later phase starts after E9.5 through EI0.S and

consists of a thin strand of cells along the lateral surface of the neural tube, which later

protrudes into the rostral sclerotome to form dorsal root ganglia and Schwarm cells. At all

stages during migration. crest cells are found on the dorsolateral pathway, in contrast to

avians.

L5. Patterning of the rhombencephalic neural crest

Neural crest cells emigrating from the rhombencephalon, populate the pharyngeal arches and

contribute to a large variety of derivatives, including cranial ganglia, the facial skeleton, the

outflow septum of the heart, the mesenchymal component of thymus and parathyroids, and

the ENS. The rhombencephalon has received much attention in recent years because the

properties of its neural crest cells suggest that they may have a leading role in the patterning

of structures at this axial level (Noden, 1988; Richman and Tickle, 1989; Noden, 1993), When

premigratory neural crest from the first arch is used to replace the premigratory crest of the

second arch of a host embryo, the grafted crest migrates in a way appropriate for its new

position and enters the second arch. Once there. however, it makes a set of structures

appropriate for its original position, i.e. first arch jaw cartilage in the second arch (Noden,

1983; Noden, 1988). Furthermore, it is able to influence the surrounding non-neural crest

derived tissues, i.e. surface ectoderm and paraxial mesoderm to fonn first arch structures

24

(beaks and jaw muscles, respectively), This suggests that neural crest cells of the

rhombencephalon acquire regional identity while still being part of the neuroepithelium, that

they carry this identity with them as they move into the pharyngeal arches. and are able to

transmit it to their surroundings (Noden. 1988).

The discovery of vertebrate Antennapedia class homeobox-containing (Hox) genes has

led to a better understanding of some of the molecular processes in vertebrate hindbrain

development (Alcarn. 1989; McGinnis and Krumlauf. 1992). Hox genes. containing a 180 bp

homeodomain, encode a group of sequence specific proteins, which are capable of binding

to the DNA. These proteins can act as transcription factors and have been implicated in many

D'"""o", of ".""'n~'100 Qr II.'"T·e "nd [Ix·e """"

"T","' ~f r~'w~~ Croup: 1. 2. 3. 4. 5. 6. 7. S. 9. 10. 11. 12. 13.

Hox ,\ (Hox I) _

@@@@@@®O@@)@O@)~ HUm," I~' I" If: 10 II; I" III \(; 111 II IJ M", .... I.. 1.11 I.~ \A ,_, I~ 1.1 1.7 III I." 1.1"

Hox B IHO\ 2)

@@@@®@@)®®OOOO Ilu",,,!1 :11 :(; ~f' ~II !~ !I; !D lE M,,,,,,, !.~ !." !.7 !.b 1.1 .:.: !..' l..l .:..<

HoxC(HoxJI

OOO@@)@O@)@)@l@)@)§ Hum," 3f: .Il>.'<: .liI .11' .11 .111 .\1-' .11; Mow,", J.4 J_' .'.I.U J ... 1.7

HoxD(Hox4) _

®O@@OOO@@@)§)@1@(§ H,mon 4G 4i\ on 4f: 4C 41) of 411 41 MO'"'' 4.1 0.1 0.: .... ' olA 4.5 4.6 4.7 4.H

Figure 5: Four mammalian Box complexes and the new and old names of the genes. The new names are a single letter (A, B, C, or D) followed by a number from 1 to 13. Genes expressed most anteriorly have the lowest numbers. The new numbers (in ovals) are shown in the order they are found along the chromosome. The most commonly used synonyms are shown below each oval. Each column of genes indicates corresponding genes in the four Box complexes based on homeodomain sequences alone. Empty ovals indicate that no gene has been detected in mice or humans. To date, all flox genes including (excluding Evx genes) appear to be transcribed in the direction shown at the bottom of the figure. Alignment with Drosophila genes is shown at the top. A strong case can be made from sequence similarity and expression pattern for the relatedness of labial (lab), proboscipedia (pb), Deformed (Dfd), and Abdominal-B (Abd-B) with groups 1, 2, 4, and 9, respectively. There are four groups in the region corresponding to sex combs reduced (scr), Antennapedia (Antp), Ultrabithorax (Ubx), and Abdominal-A (Abd-A), but more exact relations have not been detennined (indicated by brackets). Shan arrows indicate the directions of transcription of genes in ANT-C and BX-C. (From M.P. Scott, 1992, Cell 71:551-553)

25

aspects of development. Hox genes were found to be arranged in clusters along the

chromosome. During the evolution of chordates. the number of Hox gene clusters has

increased, probably through chromosomal duplication events. The acorn worm (a

hemichordate) has only one Hox cluster, two were found in amphioxus (a cephalochordate),

whereas in the lamprey (a primitive vertebrate) three Hox clusters were identified (Pendleton

et al., 1993). Vertebrates possess four clusters of Hox genes, which are known in mammals

as the Hox-A, Hox-B, Hox-C, and Hox-D (formerly called Hox-i through 4) gene clusters: Fig,

5). It is possible to identify subfamilies of up to four genes each belonging to a different

cluster (known as paralogous groups), which presumably arose from a single ancestor gene.

There is, however, not a one to one relationship between genes in the various clusters,

because of tandem duplications of genes within a cluster (Krumlauf, 1992). Genes within one

cluster display a direct linear relationship between their order along the chromosome and the

antero-posterior axial level at which they are expressed. Genes that lie most 3' within a Hox

gene cluster, have the most anterior expression restriction with cutoffs corresponding to

rhombomere boundaries (Wilkinson et al" 1989), This colinearity was first described in

Drosophila (Lewis, 1978), but was also found in vertebrates (Duboule and Dolle. 1989:

Graham et al .. 1989). Expression of a specific combination of Hox genes within each

rhombomere results in a segment restricted Hox code (Hunt et al., 1991a). This code is first

established in the neuroepithelium when still containing the neural crest and is maintained

during neural crest cell migration. This results in a Hox code in the cranial ganglia and the

pharyngeal mesenchyme. reflecting their rhombomere of origin (Lumsden et al" 1991), Upon

contact with the ectoderm, the pharyngeal mesenchyme transduces its Hox label. consistent

with the evidence that the neural crest is able to influence the development of other tissues

in the head (Hunt et al" 1991b),

In order to really prove that the expression of a particular combination of Hox genes

could be responsible for controlling the identity of a segment, the cranial Box code has been

experimentally altered. The technique of homologous recombination in embryonic stem cells

was used to produce mice lacking functional copies of genes of the Hox-A cluster. Mice

lacking the Hox-AI or the Hox-A3 gene died shortly after birth and showed defects in the

pharyngeal region (Chisaka and Capecchi. 1991: Lufkin et al" 1991: Chisaka et al" 1992),

Defects in Hox-A3 mutants occurred mainly in mesenchymal derivatives of the

rhombencephalon (Chisaka and Capecchi. 1991). whereas Hox-Ai mutants mainly displayed

abnormal development of neural structures (Lufkin et al" 1991: Chisaka et al" 1992), A

significant feature of the phenotypes of both mutants was. that while there were profound

effects on the development of structures at a particular axial level. broadly correlating with

the anterior domain of expression of the genes, other structures deriving from the same axial

level, which in some cases expressed the gene in normal development, were unaffected.

Overexpression of the Hox-A4 gene also resulted in a defective development of

26

rhombencephalic structures (Wolgemuth et aI .• 1989). These mice suffered from megacolon

caused by a non-functional ENS (Gershon and Tennyson. 1991).

Defective development of rhombencephalic structures was also observed after fetal

exposure to retinoic acid (RA), both in humans (Lammer et aI., 1985) and in a number of

animal species (Shenefelt, 1972; Fantel et aI., 1977; Kamm, 1982; Webster et aI., 1986). The

phenotype somewhat resembled the Box-A3 knock-out mice with a disturbed development of

the mesenchymal derivatives of the rhombencephalon, i.e. absence or hypoplasia of the

thymus and parathyroids and impaired outflow septation of the heart (not found in the Hox-A3

mutants). Recently, it was shown that RA is capable of altering the hindbrain Box code,

resulting in the homeotic transformation of rhombomeres 2/3 to a 4/5 identity (Marshall et

aI., 1992). After entrance into the cell and reaching the nucleus, RA can form a complex with

specific RA receptors. This RA-receptor complex acts as a transcription factor and is able to

regulate gene expression through binding to RA-responsive elements present in the promoter region of certain genes. Two cellular RA binding proteins (CRABP-I and CRABP-II), present

in the cytoplasm, are thought to modify the effect of RA. They could either function as a

shuttle to transport RA to the nucleus, or they could regulate the concentration of free

cytoplasmic RA either by steepening a RA gradient or by functioning as a sink, protecting

the nucleus from excess RA. The latter possibility is favoured by a study in F9

teratocarcinoma cells in which CRABP-I was shown to influence the metabolism of

intracellular RA (Boylan and Gudas, 1992). Additional evidence for a protective role of

CRABP-I against excess RA comes from the fact that CRABP-I has been found to be

sp~cifically expressed in tissues that seem to be sensitive to RA exposure during development

(Vaessen et aI., 1990; Maden et aI., 1991; Ruberte et aI., 1991).

1.6. In vitro studies of the neural crest

The analysis of neural crest cell migration, their state of determination and the environmental

factors involved in their differentiation, has been greatly facilitated by in vitro approaches.

Although these studies have been conducted in a number of different species. such as

amphibians (Akira and Ide, 1987; Wilson and Milos, 1987), reptiles (Hou and Takeuchi.

1992), and mammals (Ito and Takeuchi, 1984; Ito et aI .• 1988; Ito et aI .. 1993), our main

focus will be on birds.

To study factors influencing migration. neural crest cells of various axial levels were

grown in vitro. The neural tube containing the premigratory neural crest was explanted to

allow neural crest cells to emigrate from the dorsal aspect of the neural tube onto permissive

two-dimensional substrates. Neural crest cells were found to migrate avidly on planar

substrates comprised of purified extracellular matrix components such as fibronectin (Rovasio

27

et al .• 1983).laminin (Newgreen. 1984). collagen (Cohen and Konigsberg. 1975) and tenascin

(Halfter et al .• 1989). Fibronectin. found to be the most adhesive molecule. is present in high

concentration in the neural crest cell migration pathways (Newgreen and Thiery. 1980;

Duband et al .• 1986). Hyaluronate and chondroitin sulphate were found to be poor two­

dimensional migration substrates (Erickson and Turley. 1983). although addition of

hyaluronate to three-dimensional matrices of collagen or fibronectin increased neural crest cell

migration by opening up spaces between the collagen fibrils in the gel (Tucker and Erickson.

1984). Hyaluronic acid also influences adhesivity among neuroepithelial cells and could

therefore be important in the initial separation of neural crest cells from the neural tube

(Luckenbill-Edds and Carrington. 1988).

An important and still largely unanswered question concerns the developmental

potential of individual neural crest cells. Two major scenarios have been proposed to account

for the diversity of derivatives arising from the neural crest. First. the neural crest may be

composed of a homogeneous population of totipotent cells with identical developmental

potential. the fate of which thought to be completely determined by the embryonic

environment. A second possibility is that the neural crest may be comprised of a

heterogeneous mixture of predetermined cells. These committed cells would differentiate

according to an inherent program upon reaching their proper location; those in inappropriate

sites either fail to differentiate or die. Evidence in support of both schemes has been obtained.

In clonal analysis studies, some neural crest cells have been shown to contribute to multiple

phenotypes in vitro (Bronner-Fraser et al .• 1980; Sieber-Blum and Cohen. 1980; Baroffio et

al .. 1988; Sieber-Blum. 1989; Baroffio et al .• 1991; Ito and Sieber-Blum. 1991; Ito and

Sieber-Blum, 1993). Several monoclonal antibodies have been identified, however, that

specifically recognize subpopulations of early migrating neural crest cells (Ciment and

Weston. 1982; Girdlestone and Weston. 1985; Barbu et al .. 1986; Barald. 1988). suggesting heterogeneity in early neural crest cell populations.

Clonal analysis of the posterior rhombencephalic neural crest showed that it consists

of a heterogeneous mixture of both pluripotent and developmentally restricted progenitors (Ito

and Sieber-Blum, 1991). Pluripotent progenitors gave rise to sensory and serotonergic

neurons, chondrocytes, and connective tissue. smooth muscle and pigment cells. Upon entry

into the posterior pharyngeal arches. cells lost the potential to differentiate into pigment cells

and sensory neurons (Ito and Sieber-Blum. 1993). confirming earlier observations by Ciment

et al. (Ciment and Weston. 1983; Ciment and Weston. 1985). The developmental potential

to form serotonergic neurons, which may constitute precursors for enteric neurons. also

decreased considerably upon entry into the pharyngeal arches (Ito and Sieber-Blum. 1993).

These results show that although the posterior rhombencephalic neural crest contains

pluripotent precursors which can give rise to serotonergic neurons, most precursors loose this

capacity upon entry into the pharyngeal arches, resulting in a high percentage of clones which

28

UF BDNF

SCF bFGF

Neural crest

8

Sympa!t1etlc ~ nouron

F,. sympathotic

neuron

Enteric neuron

Figure 6: Actions of growth factors on the differentiation of neural crest cells in vertebrates. A single growth factor can regulate the differentiation of many distinct cell types at different stages of embryonic development. Conversely, members of different classes of growth factors can act on the same group of progenitor cells to induce similar developmental programs. (b)FGF, (basic) fibroblast growth factor; SCF, stem cell factor; LIF, leukemia inhibitory factor; BDNF, brain derived neurotrophic factor; NGF, nerve growth factor; SIF cell, small intensely fluorescent cell.

only consist of ectomesenchymal cell types.

The embryonic microenvironment may play an important role in the emergence of

phenotypic diversity. Certain factors could act to promote the survival of selected

subpopulations of fully determined progenitors, while others may direct partly committed

precursors toward a specific developmental fate (Howard and Bronner-Fraser, 1985; Howard,

1986; Ziller et al .. 1987; Barald, 1989). Differentiation can be influenced both through direct

contact with tissues or extracellular matrix and through diffusible factors (Fig. 6).

1.7. Clinical disorders related to the neural crest

Of all the described human congenital malformations (defined as structural defects present at

birth), about one third entails structures related to the neural crest (Table 1). In 1974 Bolande

(1974) introduced the term neurocristopathies for certain tumors, such as pheochromocytoma,

neuroblastoma, and neurofibromatosis type 1 and 2, occurring either isolated or in

combination. In this way, he attempted to delineate a common pathogenetic denominator for

a heterogeneous group of disorders, which would be aberrant neural crest development. Since

then. the term neurocristopathy has been used for any malformation concerning structures that

receive some contribution from the neural crest.

29

Table 2: Pharyngeal arch related syndromes and their characteristic defects

Syndromes DiGeorge vep' Goldberg-Sphrintzen

Defects

Related to Micrognathia Micrognathia Micrognathia arches IJII (Fishmouth)

Related to Ear defects Ear defects Ear defects arches Vascular defects Vascular defects Vascular defects IlUIV Heart defects Heart defects Heart defects

Thymus aplasia Thyroid defects Parathyroid defects

Others Vertebral defects Cerebral defects3 Cerebral defects] Cerebral defects}

Deafness}

Prominent nose Broad nasal root

I) Velocardiofacial or Sphrintzen syndrome 2) Familial heart disease J) not always found ~) Disorders of the enteric nervous system

Deafness3

ENS defccts4

Prominent nose Broad nasal root

VACTERL Goldenhar

Micrognathia Uni- or bilateral facial hypoplasia Eye defects

lvficrotia Vascular defects Heart defects Heart defects

Vertebral defects Vertebral defects

Anal atresia Tracheal fistula Esophageal fistula Renal defects Limb defects

CHARGE

Micrognathia

Ear defects

Heart defects

Cerebral defects Deafness

Choanal atresia Coloboma Genital defects

FHD2

Hearl defects

As has become apparent in this chapter. the neural crest contributes to a vast amount

of structures throughout the body. Most of these structures, however, are not entirely neural

crest derived. but involve interactions with other germ-layers. Ear development. for example.

involves interactions between ectoderm (flrst pharyngeal cleft and otic placode). mesoderm

(mostly neural crest cells from pharyngeal arches 1-4) and endoderm (first pharyngeal pouch),

each giving rise to a speciflc structure of ear (Larsen, 1993). Furthermore, complex

malformations often occur in syndromes. with a characteristic set of malformations occurring

in the same patient. Waardenburg type I syndrome. for example. is characterized by

congenital deafness. pigment abnormalities (white forelock) and various other anatomical

changes (Omenn and McKusick, 1979; Badner and Chakravarti, 1990). It is caused by a

mutation in the PAX-3 gene, which, in mice, was found to be expressed in the dorsal part of

the neural tube, including the neural crest (Baldwin et al., 1992; Gruss and Walther, 1992;

Tassabehji et al .. 1992). Although part of the malformations in this syndrome could be caused

by a genetic defect in the neural crest. not all of the malformations can be explained by

defective neural crest development.

Many syndromes which are considered to be neurocristopathies. like Waardenburg type

I syndrome. involve structures in the head and neck region, which are related to the

pharyngeal arches. Syndromes related to the third and fourth pharyngeal arches may entail

malformations of the outflow tract of the heart, the thymus, the parathyroids, and the ENS,

often combined with craniofacial dysmorphologies. Well-known examples are the DiGeorge,

Goldenhar, Velocardiofacial and Goldberg-Sphrintzen syndromes as well as the CHARGE

association (Table 2). Careful examination of Table 2 shows that there is considerable overlap

in the malformations found in these syndromes. This shows the difficulty of syndrome

delineation and stresses the importance of careful examination of patients.

1.8. Conclusions

Comparing neural crest development in various vertebrate species shows that, although there

are species specific differences in both the timing and pathways of neural crest cell migration

and differentiation, the basic principles are very similar. We chose the chicken embryo as a

model system in the experimental work, because, like amphibians, it is easily amenable to

experimental manipulation. Amphibians. however, differ from birds and mammals in the

extend of cephalization, which can be illustrated by the evolution of the skull (Augier, 1931;

Couly et al., 1993). Primitive vertebrates, such as the agnatha. have a small archiskull in

which no vertebrae participate, whereas higher vertebrates. such as birds and mammals, have

a neoskull in which 5 somites participate in the occipital bone complex. Amphibians have an

intermediate skull type, a paleoskull, in which 3 somites are incorporated. This difference in

31

skull type has consequences for the neural crest of somites 4 and 5, which belongs to the

cranial crest in birds and mammals, but should be considered trunk crest in amphibians. The

chicken embryo is also well suited to study the role of various growth factors on neural crest

cell differentiation in vitro.

Although most studies support the similarity of neural crest development in birds and

mammals, one should remain careful in the extrapolation of data. This can be illustrated by

comparing certain pharyngeal arch derivatives. In birds, the cartilages of the first pharyngeal

arch give rise to the jaw joint. Among the immediate ancestors of mammals, however, a

second, novel jaw articulation developed. As a result, the cartilage that formed the jaw joint

in non-mammals, shifted to the middle ear and joined with the preexisting stapes (derived

from the cartilage of the second pharyngeal arch) to form the unique three-ossicle auditory

mechanism of mammals (Larsen, 1993),

1.9. References

Akum, M. (1989) Hox and HOM: homologous gene elusters in inseets:md vertebrates. Cell 5:347-349.

Akira, E., and Ide, H. (1987) Differentiation of neural crest celis of Xenopus laevis in clonal culture. Pigment Cell Res. 1:28-36.

Artinger. KB .. and Bronner-Fraser, M. (1992a) Notochord grafts do not suppress formation of neural crest cells or commissural neurons. Development 116:877-886.

Artinger. KB .. :md Bronner-Fraser, M, (1992b) Partial restriction in the developmental potential of late cmigrating avi:m neuroJ crest cells. Dev. BioI. 149:149-157.

Augier. M. (1931) Squelette cepJw1ique. In: Squeletle cephalique. Poirier:md Charpys (cd). Masson et Cic: Paris.

Badner, J.A .. and Chakravarti. A. (1990) Waardenburg syndrome and Hirsehsprung disease: evidence for pleiotropic effects of a single dominant gene, Am, J. Med, Genet 35:100-104.

Baldwin. C.T .. Hoth, C.F .. Amos. J.A" da-Silva.. E.O .. and Milunsky, A. (1992) An exonic mutation in the HuP2 paired domain gene causes Waardenburg's syndrome. Nature 355:637-638.

Bancroft. M .. and Bellairs. R. (1976) The neural crest of the trunk region of the chick embryo studied by SEM :md TEM, Zoon 4:73-85,

Barald, K (1988) Monoclonal antibodies made to chick mesencephalic neural crest cells and to ciliary ganglion neurons identify a common antigen on the neurons and a neural crest subpopulation. J. Neurosci. Res. 21:107-118.

Barald. KF. (1989) Culture conditions affect the cholinergic development of an isolated subpopuiation of chiek mesencephalic neural Crest cells. Dev. BioI. 135:349-366.

Barbu. M .. Ziller, C .. Rong. P.M .. and LeDouarin, N.M. (1986) Heterogeneity in migrating neural crest cells revealed by a monoclonal :mtibody. J. Neurosci. 6:2215-2225.

Baroffio. A" Dupin, E .• and LcDouarin. N.M, (1988) Clone-forming ability and differentiation potcntial of migratory neural crest cclls. Proc. Nat!. Acad. Sci. USA 85:5325-5329.

Baroffio, A .. Dupin. E .. and LeDouann. N.M. (1991) Common precursors for neural and mesectodermal derivatives in the cephalic neural crest. Development 112:301-305,

Bayliss. W.M .. and Starling, E.H. (1899) The movcmcnts and innervation of the small intestine. J. Physiol. 24:99-143.

Bayliss. W.M .. and Starling. E.H. (1900) The movements and innervation of the large intestine. J. Physio!' 26:107-118.

32

Bockman. D.E .• and Kirby. M.L. (1984) Dependence of thymus development on derivatives of the neural crest. Science 223:498·500.

Bolande. R.P. (1974) The neuroeristopathies. A unifying concept of disease arising in neural crest maldevelopment. Hum. Pathol. 5:409429.

Boylan. IF,, and Gudas. L.J. (1992) The level of CRABp·! expression influenccs thc amounts and types of all·tr:ms-retinoic acid metabolitcs in F9 teratocarcinoma stem cells. J. BioI. Chern. 267:21486-21491.

Bronner-Frascr. M .. and Cohen. A.M. (1980) Analysis of the neural crest vcntral pathway using injccted traccr cells. Dev. Bioi. 77:130-141.

Bronner-Fraser. M .. Sieber-Blum. M" and Cohen. A.M. (1980) Clonal analysis of the avian neural crest: migration and maturation of mixed neural crest clones injected into host chicken embryos. J. Compo Neurol. 193:423-434.

Bronner-Fraser. M. (1985) Alterations in neural crest migration by a monoclonal antobody that affects cell adhesion. J. Cell BioI. 101:610·617.

Bronner-Fmser. M. (1986) An antibody to a receptor for fibronectin and laminin perturbs cranial neural crest development in vivo. Dev. BioI. 117:528-536.

Bronner-Praser. M. (1987) Perturbation of cranial neural crest migration by the HNK-J antibody. Dev. BioI. 123:321-331.

Bronner-Praser. M .. and Fraser. S.E. (1988) Cell lineage analysis reveals multipotency of some avian neural crest cells. Nature 335:161-164.

Bronner-Fraser. M" and Lallier. T. (1988) A monoclonal antibody against a laminin-heparan sulfate proteoglycan complex perturbs cranial neural crest migmtion in vivo. J. Cell BioI. 106:1321·1330.

Bronner-Fraser. M .. Stem. C.D .. and Fraser. S. (1991) Analysis of neural crest cell lineage and migration. J. Craniofac. Genet. Dev. Bioi. J 1 :214-222.

Bronner-Fraser. M" Wolf. J.J .. and Murray. B.A. (1992) Effects of antibodies against N-Cadherin and N-CAM on the cmnial ncural crest and neural tube. Dcv. BioI. 153:291-301.

Chan. W.Y .. and Lee. K.K.H. (1992) The incorpomtion and dispersion of cells and latex beads on microinjection into the amniotic cavity of the mouse embryo at the early-somite stage. Anat. Embryo!. 185:225-238.

Chibon. P. (1964) Analyse par In methode de marquage nucieaire a In thymidine tritiec des. derives de la crete neurale ccphalique chez I'Urodcle Pleurodeles waltlii. C. r. Acad. Sci. 259:3624-3627.

Chibon. P. (1966) Analyse experimentale de ]a rcgionalis:ltion et des capacitcs morphogcnctiques de la crete chez J'amphibicn urodcle Pleurodeles Waltlii Michah. MCm. Soc. zool. Fr. 36:1-107.

Chisaka. 0 .. and Capecchi. M.R. (1991) Regionally restricted developmental defects resulting from targeted disruption of the mouse homeobox gene Hox-1.5. Nature 350:473479.

Chisako.. 0 .. Musci. T.S .. and Capecchi. M.R. (1992) Developmental defects of the car. cranial nerves and hindbrain resulting from t::trgeted disruption of the mouse homeobox gene Hox-1.6. Nature 355:516·520.

Ciment. G .. and Weston. J.A. (1982) Early appearance in neural crest and crest-derived cells of an antigenic detcnninant present in avian neurons. Dev. BioI. 93:355-367.

Ciment. G .. and Weston. J.A. (1983) Enteric neurogcnesis by neural crest-derived bmnchial arch mesenchymal cells. Nature 305:424-427.

Ciment. G .. and Weston. J.A. (1985) Segregation of developmental abilities in neural crest-derived celIs: identification of partially restricted intermediate cell types in the branchial arches of avian embryos. Dev. BioI. 111 :73-83.

Cohen. A.M .. and Konigsberg. I.R. (1975) A cIon:l.! approach to the problem of neural crest determination. Dev. BioI. 46:262-280.

Collazo. A .. Bronner·Fraser. M .. and Fraser. S.E. (1993) Vital dye labelling of Xenopus laevis trunk neural erest reveals multipotency and novel pathways of migration. Development 118:363-376.

Cooke. H,J. (1986) Neurobiology of the intestinal mucosa. Gastroenterology 90:1057-1081.

33

Couly. G.F .• Coltey. P.M., and LeDouarin, N.M. (1993) The triple origin of skull in higher vertebrates: a study in quail-chick chimeras, Development 117:409-429.

Detwiler, S.R. (1937) Observations upon the migration of neural crest cells and upon the development of the spinal ganglia and vertebral arches in Amblystoma, Am. J. Anat. 61:63·94.

Duband. J.L .• Rocher. S .. Chen. W,T .. Yamada. KM .. and Thiery, J.P. (1986) Cell adhesion and migration in the early vertebrate embryo: Location and possible role of the putative fibroneetin receptor complex. J. Cell BioI. 102:160-178.

Duboule. 0 .. and 00116, P. (1989) The structural and functional orgo.nization of the murine HOX gene family resembles that of Drosophila homeotic genes. EMBO J, 8:1497-1505.

Eppcrlein, H .. Halfter, W .. and Tucker, RP, (1988) The distribution of fibroneetin and tenascin along migratory pathways of the neural crest in the trunk of amphibian embryos, Development 103:743-756,

Erickson, C.A .. and Turley, E.A (1983) Substrata formed by combinations of extracellular m:ltrix components alter neural crest cell motility in vitro, J. Cell Sci. 61:299-323.

Erickson, C.A .. and Weston, J.A, (1983) A SEM analysis of neural crest migration in the mouse. J. Embryo!. expo MorphoL 74:97-118.

Erickson. C,A .. Loring, IF .. and Lester, S.M. (1989) Migratory pathways of HNK-l-immunoreactive neural crest cells in the rat embryo. Dev. Bio!. 134:112-118.

Erickson, C.A .. Duong, T.D .. and Tosney, KW. (1992) Descriptive and experimental analysis of the dispersion of neural crest cclls along the dorsolateral path and their entry into ectoderm in the chick embryo. Dev. BioI. 151:251-272,

Ewart. W.R. (1985) Sensation in the gastrointestinal tract. Compo Biochem. Physio!. 82A:489-493.

FanteL A,G .• Shepard. T.H .. Newell-Morris. L.N .. and Moffett. B.c. (1977) Teratogenic effects of retinoic acid in pigtail monkeys (Macaca nemestrina). Teratology 15:65-72.

Felten, D.L .. Felten. S.Y .. Bellinger. D.L .. Carlson. S.L .. Ackerman. K.D .. Madden. KS .. Olschowski, JA. and Livnat. S. (1988) Noradrenergic sympathetic neural interactions with the immune system: structure and function. ImmunoL Rev. 100:225-260.

Gans. C .. and Northcutt. G. (1983) Neural crest and the origin of vertebrates: A new head. Science 220:268-274.

Gershon. M.D. (1981) The enteric nervous system. Ann. Rev. Neurosci. 4:227·271.

Gershon. M.D .. and Tennyson. V.M. (1991) Microcnvironmcntaljactors in the nonnal and abnormal development 01 the enteric nervous system. Wiley-Liss. Inc.: New York. 257-276.

Girdlestonc, J.. and Weston. J.A. (1985) Identification of early neuronal subpopulations in avian neural crest cell cultures. Dev. BioI. 109:274-287.

Goldstcin. R.S .. and Kaleheim. C. (1991) Normal segmentation and size of the primary sympathetic ganglia depend upon the alternation of rostrocaudal properties of the somites. Development 112:327-334.

Graham, A, Papalopulu. N .. and Krumlauf. R. (1989) The murine and Drosophila homeobox elusters have common features of organization and expression. Cell 5:367-378.

Graham. A .• Heyman. I.. and Lumsden, A. (1993) Even-numbered rhombomeres control the apoptotic elimination of neural crest cells from odd~numbered rhombomeres in the chick hindbrain. Development 119:233-245.

Gruss. P .. and Walther. C. (1992) Pax in development. Cell 69:719·722.

Halfter. W .. Chiquet-Ehrismann. Roo and Tucker. R.P. (1989) The effect of tenascin and embryonic basal lamina on the behavior and morphology of neural crest cells in vitro. Dev. BioI. 132:14·25.

Hall. B. (1992) Evolutionary developmental biology. Chapman and Hall: London.

Hamburger, V .. and Hamilton. H.L. (1951) A series of normal stages in the development of the chick embryo. J. Morpho\, 88:49-67.

Horstadius, S .. and Sellman. S. (1946) Experimentelle untersuchungen tiber die Determination des Knorpcligen Kopfskelettes bei Urodelen. Nova Acta R. Soc. Scient. Upsa!. 13:1-170.

34

Hou. L.. and Takeuchi. T. (1992) Differentiation of reptilian neural crest cells in vitro. In Vitro CelL Dev. BioI. 28A:348-354.

Howard. M.J., and Bronner-Fraser. M. (1985) The influence of neural tube-derived factors on differentiation of neural crest cells in vitro. I. Histochemical study on the appearance of adrenergic cells. J. Neurosci. 5:3302-3309.

Howard. MJ. (1986) Neural crest cell diffcrentiation in vitro: factors affecting expression of the adrenergic phenotype. Prog. Clin. Bio!. Res. 217B:267-272.

Hunt. P .• Gulisano, M., Cook, M" Sham, M .. Faiella, A,. Wilkinson, D .. Boncinelli, E" and Krumlauf. R. (199la) A distinct Hox code for the branchial region of the vertebrate head. Nature 353:861-864.

Hunt. P .. Wilkinson. D" and Krumlauf. R. (l991b) Patterning the vertebrate head: murine Hox 2 genes mark distinct subpopulations of premigratory and migrating neural crest. Development 11:43-51.

Ito, K" and Takeuchi. T. (1984) The differentiation in vitro of the neural crest cells of the mouse embryo. J. Embryo!. Exp. Morph. 84:49-62.

Ito. K" Morita. T., and Takeuchi. T. (1988) Neuronal differentiation of mouse neural crest cells in vitro. Cell Struct. Funct. 13:267-270.

Ito. K" and Sieber-Blum, M. (I 991) In vitro clonal analysis of quail cardiac neural crest development. Dev. BioI. 148:95-106.

Ito. K. Morita, T" and Sieber-Blum. M. (1993) In vitro clonal analysis of mouse neural crest dcvelopment. Dev. BioI. 157:517-525.

Ito. K" and Sieber-Blum. M. (1993) Pluripotent and developmentally restricted neural-crest-derived cells in posterior visceral arches. Dev. Bio!. 156:191-200.

Jacobson. A.G. (1981) Morphogenesis of the neural plate and tube. T.G. Conelly. L.L. Brinkley. and B.M. Carlsons (ed). Raven Press: New York. 233-263.

Jaenisch. R. (1985) Mammalian neural crest cells participate in normal embryonic development on microinjection into post-implantation mouse embryos. Nature 318:181-183.

Jeffs, p" Jaques. K, and Osmond. M. (1992) Cell death in cranial neural crest development. Anat. Embryo!. 185:583-588.

Jeffs, P" and Osmond. M. (1992) A segmented pattern of cell death during development of the chick embryo. Annt Embryo!. 185:589-598.

Jesse!. T.M" Bovolenta, P" Placzek. M" Tessier-Lavigne. M" and Dodd, J. (1989) Polarity andpatteming in the neural tube: the origin andfunction of the floor plate. D. Evered and J. Marshs (cd). Wiley and Sons: Chichester. 255-280.

Johnston, M.e. (1966) A radioautographic study of the migration and fate of cranial neural crest cells in the chick embryo. Anat. Rec. 156:143-156.

Kalchcim. C. and Teillet. M. (1989) Consequences of somite manipulation on the pattern of dorsal root ganglion development. Development 106:85-93.

Kamm, U. (1982) Toxlcology. carcinogenicity. and teratogenicity of some orally administered retinoids. J. Am. ACJ.d. Dermato!. 6:652-659.

Keynes. RJ .. and Stem. C.D. (1984) Segmentation in the vertebrate nervous system. Nature 310:786-789.

Keynes, R .. and Lumsden. A. (1990) Segmentation and the origin of regional diversity in the vertebrate central nervous system. Neuron 4:1-9.

Kirby, M.L .. Gale. T.F .. and Stewart, D.E. (1983) Neural crest cells contribute to normal aorticopulmonary septation. Science 220:1059-1061.

Kirby, M.L., and Stewart. D.E. (1983) Neural crest origin of cardiac ganglion cells in the chick embryo: IdentifiCJ.tion and extirpation. Dev. BioI. 97:433-443.

Krotoski. D .. Domingo. C" and Bronner-Fraser. M. (1986) Distribution of a putative cell surface receptor for fibronectin and lamirun in the avian embryo. J. Cell BioI. 103:1061-1072.

Krotoski. D.M" Fraser, S.B .. and Bronner-Fraser. M. (1988) Mapping of neural crest pathways in Xenopus laevis using inter­and intra-specific cell markers. Dev. BioI. 127:119-132.

35

Krumlauf, R. (1992) Evolution of the vertebrate Hox homeobox genes. Bioessays 1:245-252.

Kruse, J .• Mailhammer, R., Wernecke, H .. Faissncr, A" Sommer, 1., Goridis, C., and Schachner, M, (1984) Neural cell adhesion molecules and myelin·associated glycoprotein share a common carbohydrate moiety recognized by monoclonal antibodies L2 and HNK-I. Nature 311:153·155.

Kuntz, A. (1963) The autonomic nervous system. Lea & Febiger: Philadelphia.

Kuratani, S,c., and Kirby, M.L. (1991) Initial migration and distribution of the cardiac neural crest in the avian embryo: An introduction to the concept of the circumpharyngeal crest. Am. 1. Anat. 191:215-227.

Lallier, T" and Bronner-Fraser, M. (1988) A spatial and temporal analysis of dorsal root and sympathetic ganglion formation in the avian embryo, Dev, Bio!. 127:99-112,

L:lID.mer, E.J.. Chen. D.T., Hoar, R,M .. Agnish, N,D .. Benke, PJ., Braun, J,T .. Curry, C,J .. Fcmhoff, P.M .. Grix, A.W" Lott. LT., Richard, J.M .. and Sun, S.c. (1985) Retinoic acid embryopathy, N. Eng!. J. Med. 313:837·841.

Langley, J.N. (1921) The autonomic ner,Jous system. W. Heffer: Cambridge,

Larsen, W,J, (1993) Human embryology. W.R, Schmitt, M. Otway, and E, Bowman-schulman (cd.). 1993, Churchill Livingstone: New York.

LeDouarin, N.M., and Teillet. M. (1973) The migration of neural crest cells to the wall of the digestive tract in avian embryo. J. Embryo!. expo Morpho!. 30:31-48.

LcDouarin. NoM. (1982) The neural crest. Cambridge Univ. Press: Cambridge.

Lewis, E. (1978) A gene complex controlling segmentation in Drosophila, Nature 27:565-570.

Lim, T,M .. Lunn, E.R., Keynes, R,J .. and Stem, C.D. (1987) The differing effects of occipital and trunk somites on neural development in the chick embryo. Development 100:525-533,

UHberg, J., Ahlfors, K.. and Ftlllstr6m, C. (1980) Neural crest cell migration in relation to cxtracellular matrix organization in the embryonic axolotl trunk, Dev. BioI. 75:148-167.

L6fbcrg, J., and Nynas·MeCoy, A. (1981) Epidermal stimulation of initial neural crest cdl migration in the amphibian embryo, Basel.

Lofberg, J" Nyntis-MeCoy, A .. Olsson. C" J6nsson, L" and Perris, R. (1985) Stimulation of initial neural crest cell migration in the axolotl embryo by tissue grafts and extracellular matrix transplanted on microcarriers. Dev. BioI. 107:442-459,

LMberg. J.. Perris, R .. and Epperlcin, H.H, (1989) Timing in the regulation of neural crest cell migration: retarded "maturation" of regional extracellular matrix inhibits pigment Cell migration in embryos of the white axOlotl mutant. Dev. Bio!. 131:168-181.

Luckenbill-Edds. L .. and Carrington, J.L. (1988) Effect of hyalUronic acid on the emergence of neural crest cells from the neural tube of the quail. Cotumix cotumix japonica. Cell Tissue Res. 252:573-579,

Lufkin, T., Dierich, A., LeMeur, M .. Mark, M., and Chambon, p, (1991) Disruption of the Hox-1.6 homeobox gene results in defects in a region corresponding to its rostral domain of expression, Cell 66:1105·1119,

Lumsden, A .. Sprawson, N .. and Graham, A, (1991) Segmental origin and migration of neural crest cclls in the hindbrain region of the chick embryo. Development 113:I281~1291.

MacMillan, G.J. (1976) Melanoblast tissue interactions and the development of pigment patterns in Xenopus larvae. J. Embryo!. expo Morpho!. 35:463-484.

Maden, M .. Hunt, p" Eriksson, U .. Kuroiwa, A .• Krumlauf, R" and Summerbell, D. (1991) Retinoic acid·binding protein, rhombomeres and the neural crest. Development 111:35-44.

Maisey. J,G. (1986) Heads and tails: a chordate phylogeny. Cladistics 2:201-256.

Man;hall. H" Nonchev, S .. Sham. M,H .. Muchamorc, 1., Lumsden, A .. and Krumlauf. R. (1992) Retinoic acid alters hindbrain Hox code and induces transformation of rhombomeres 2/3 into a 4/5 identity, Nature 360:737~741.

Matthew, W.D .. and Patterson, P,H. (1983) The production of a monoclonal antibOdy that blocks the action of a neurite Outgrowth.promoting f::lctoL Cold Spring Harbor Symp. Quant. BioI. 48:625·631.

36

McGinnis. W .• and Krumlauf. R. (1992) Homeobox genes::md axial patterning. Cell 68:283-302.

Miyagawa-Tomita. S .• Waldo. K .. Tomita. H .. and Kirby. M.L. (1991) Temporospatial study of the migration::md distribution of e:lfdi:J.e neural crest in quail-chiCk chimeras. Am. J. Anat. 192:79-88.

Moury. J.D .. and Jacobson. A.G. (1989) Neural fold fonnation at newly created boundaries between neural plate and epidcnnis in the axolotl. Dev. BioI. 133:44-57.

Moury. J.D .. ::md Jacobson. A.G. (1990) The origins of neural crest cells in the axolotl. Dev. Biol. 141:243-253.

Newgreen. D .. and Thiery. J.P. (1980) Fibroneetin in early avian embryos: synthesis and distribution ::tlong migration pathways of neural crest cells. Cell Tissue Res. 211:269-291.

Newgreen. D .. and Gibbins. I. (1982) Factors controlling the time of onset of the migration of neural crest cells in the fowl embryo. Cell Tiss. Res. 224:145-160.

Newgreen. D. (1984) Spreading of explants of embryonic chick mesenchyme and cpithelia on fibror.ectin and laminin. Cell Tissue Res. 236:265-277.

Newgreen. D.F .. Schecl. M .. and Kastner. V. (1986) Morphogenesis of sclerotome and neural crest in avi::m embryos: in vivo and in vitro studies on the role of the notochordal extracellul:lf matrix. Cell Tiss. Res. 244:299-313.

Nichols. D.H. (1981) Neural crest cell fonnation in the head of the mouse embryo as observed using a new histological teChnique. J. Embryo!. expo Morphol. 64:105-120.

Nichols, D.H. (1986) Formation and distribution of neural crest mesenchyme to the first pharyngeal arch of the mouse embryo. Am. J. Anal. 176:221-232.

Noden. D.M. (1975) An analysis of the migratory behavior of avi::m cephalic neural crest cells. Dev. BioI. 42:106-130.

Noden. D.M. (1983) The role of the neural crest in patterning of avian cranial skeletal. connective and muscle tissues. Dev. Bio!. 96:144-165.

Noden. D.M. (1988) Interactions and fates of avian craniofacial mesenchyme. Development 103 Suppl.:12l-140.

Noden. D.M. (1993) Spatial integration among cells fonning the cranial peripheral nervous system. J. Neurobiol. 24:248-261.

Norris. W.E .. Stern. C.D .. ::md Keynes. R.I. (1989) Molecular differences between the rostral and caudal halves of the selerotome in the chick embryo. Development 105:541-548.

Northcutt. G .. and Gans. C. (1983) The genesis of neural crest and epidermal plaeodes: a reinterpretation of vertebrate origins. Q. Rcv. Bioi. 58:1-28.

Omenn. G.S .. and McKusick. V.A. (1979) The association of Waardenburg syndrome and Hirschsprung megacolon. Am. J. Med. Genet. 3:217-223.

Pendleton. J.W .. Nagai. B.K .. Murtha. M.T .. ::md Ruddle. F.H. (1993) Expansion of the Hox gene family and the evolution of chordates. Proc. Nat!. Acad. Sci. 90:6300-6304.

Peters-van der Sanden, M.J.H .. Kirby. M.L .• Gittenberger-de Groot. A .. Tibbocl. D .. Mulder. M.P .. and Meijers. C. (1993) Ablation of various regions within the avian neural crest has diffcrential effects on ganglion formation in the fore-. mid-. and hindgut. Dev. Dyn. 196:183-194.

Pettway. Z .. Guillory. G .. and Bronner-Fraser. M. (1990) Absence of neural crest cells from the region surrounding implanted notoehords in situ. Dev. BioI. 142:335-345.

Poole. TJ .. and Thiery, J.P. (1986) Antibodies wui synthetic peptides that block cell-fibronectin adhesion arrest neural crest migralion in vivo. H. Slavkins (cd). Alan L. Liss: New York. 235·238.

Ranscht. B .. and Bronner-Fraser. M. (1991) T-cadherin expression alternates with migrating neural crest cells in the trunk of the avian embryo. Development 111: 1 5-22.

Raven. C.P. (1931) Die cigentilmliche Bildungswcise des Ncuralrohrs beim Axolotl und die Lage des Ganglienlcistenmatcrials. Anat. Anz. 71: 161-166.

Richman. J.. and Tickle. C. (1989) Epithelia arc interchangeable between facial primordia of chick embryos and morphogenesis is controlled by the mesenchyme. Dev. BioI. 13:201-210.

37

Rickmann. M .• Fawcett. J .. and Keynes. R.J. (1985) The migration of neural crest cells and the growth of motor axons through the rostral half of the chick somite. J. Embryo!. expo Morph. 90:437-455.

Roman. C .. and Gonclla. J. (1981) Extrinsic control of digestive tract motility in physiology of the gastrointestinal tract. Raven Press: New York.

Rovasio. R.A .. Delouvee. A. Yamada. K.M .. Timpl. R., and Thiery. J.P. (1983) Neural crest cell migration; Requirements for exogenous fibronectin and high eell density. J. Cell BioI. 96:462-473.

Rubertc. B .. Dolle. p .. Chambon. P., and Morris~Kay, G. (1991) Retinoic acid receptors and cellular retinoid binding proteins II. Their differential pattern of transcription during early morphogenesis in mouse embryos. Development III :45-60.

Schroeder. T.E. (1970) Neurulation in Xenopus laevis. An analysis and model based upon light and electron microscopy. J. Embryo!. expo Morphol. 23:427-462.

Sechrist. J .• Serbedzija. G.N .. Scherson. T .• Fraser. S.E .• and Bronner-Fraser. M. (1993) Segmental migration of the hindbrain neural crest docs not arise from its segmental generation. Development 118:691"703.

Serbedzija. G.N .• Bronner"Fraser. M .. and Fraser. S.E. (1989) A vim! dye analysis of the timing and pathways of avian trunk neural crest cell migration. Development 106:809-816.

Serbedzija. G.N .. Fraser. S.B .. and Bronner-Fraser. M. (1990) Pathways of trunk neuml crest ccll migration in the mouse embryo as revealed by vital dye labelling. Development 108;605-612.

Scrbedzija. G.N .. Burgan. S .. Fraser. S.B .. and Bronner-Fraser. M. (1991) Vital dye labeling demonstrates a sacral neural crest contribution to the enteric nervous system of chick and mouse embryos. Development 111;857"867.

Serbedzija. G.N .. Bronner-Fraser. M .. and Fraser. S.E. (1992) Vital dye analysis of craniu! neural crest cell migration in the mouse embryo. Development 116:297-307.

Seufert. D.W .. and Hall. B.K. (1990) Tissue interactions involving craniul neuml crest in cartilage fonnation in Xenopus laevis (Daudin). Cell Diff. Dev. 32:153-166.

Shenefelt. R.E. (1972) Morphogenesis of malfonnations in hamsters caused by retinoie acid: Relation to dose and stage at treatment. Teratology 5:103-118.

Sieber-Blum. M .. and Cohen. A.M. (1980) Clonal analysis of qu:lil neural crcst cells: they arc pluripotent and differentiate in vitro in the absence of non-crest cells. Dev. Bioi. 80;96-106.

Sieber~Blum. M. (1989) Commitment of neural crest cel1.~ to the sensory neuron lineage. Science 243:1608"1611.

Smits"van Prooije. AE .. Venneij-Kcers. C .. Poelm:mn. R.E .. Mentink. M.M .. and Dubbeldam. J.A. (1988) The formation of mesodenn:md mescctodcnn in 5- to 41-somite rat cmbryos cultured in vitro. using WGA-Au as a marker. Anat. Embryo!. 177:245-256.

Stem. C.D .. and Keynes. R.I. (1987) Interactions between somite cells: the formation and maintenance of segment boundaries in the chick embryo. Development 99:261-272.

Stern. CD .. Norris. W.E .. Bronner-Fraser. M .. Carlson. G.I .. Faissner. A .. Keynes. R.I .. and Schachncr. M. (1989) J1/cenascin-related molecules are not responsible for the segmented pattern of neural crest cells or motor axons in the chick embryo. Development 107;309-319.

Stern. CD .. Artinger. K.B .. and Bronner-Fraser. M. (l99la) Tissue interactions affecting the migration and differentiation of neural crest cells in the chick embryo. Development 113:207-216.

Stern, CD .. Jaques. K.F .. Lim. T .. Fraser. S.E .. and Keynes. R.I. (1991b) Segmcntallincage restrictions in the chick embryo spinal cord depend on the adjacent semites. Development 113:239-244.

Stocker. K.M .. Brown. AM.C .. and eiment. G. (1993) Gene transfer of LacZ into avi:m neural tube and neural crest cells by retroviral infection of grafted embryonic tissues. J. Neurosci. Res. 34;135-145.

Stone. L.S. (1922) Experiments on the development of the emnial ganglia and the lateral line sense organs in Amblystoma punctatum. J. expo ZooL 35:421-496.

Stone. L.S. (1926) Further experiments on the extirpation and transplantation of mesectodenn in Amblystoma puncta/um. J. expo ZooL 44:95-131.

38

Stone, L.S. (1929) Experiments showing the role of migrating neural crest (mesectoderm) in the formation of head skeleton :md loose connective tissue in Rana palustris. Wilhelm Roux Arch. EntwMeeh. Org. 118:40-77.

Tan. S.S .. :md Moms*K:J.y. G.M. (1985) The development and distribution of the cranial neural crest in the rat embryo. Cell Tiss. Res. 240:403*416.

Tan. S.S .. and Moms*Kay. G.M. (1986) Analysis of cranial neural crest cell migration and early fates in postimpl:lntation rat chimeras. J. Embryol. expo Morphol. 98:21~58.

Tassnbehji. M .. Read. A.P .. Newton. V.E .. Harris. R .. Balling. R .. Gruss. P .. :md Strachan. T. (1992) Waardenburg's syndrome patients have mutations in the human homologue of the Pax-3 paired box gene. Nature 355:635*636.

Teillet, M .. Kalcheim. C .. and LeDouarin. N.M. (1987) Formation of the dorsal root ganglia in the avian embryo: Segmental origin:md migratory behavior of neural crest progenitor cells. Dev. BioI. 120:329-347.

Tosney. K,W. (1978) The early migr:ttion of neural crest cells in the trunk region of the avi:m embryo: An electron microscopic study. Dev. BioI. 62:317-333.

Trendc1enburg. P. (1917) Physioiogische und pharmakologische Versoehc ubcr die Dunnd:lI1Tl peristaltick. Arch. Exp. Pathol. Pharmaeol. 81:55-129.

Tucker. R.P .. and Erickson. CA. (1984) Morphology and behavior of quail neural crest cells in artificial three*dimensional extmcelluiar matrices. Dev. BioI. 104:390-405.

Vaage. S. (1969) The segmentation of the primitive neumi tube in chick embryos (Gallus domesticus). Adv. Anat. Embryol. Cell BioI. 41: 1-88.

Vaessen. M .. Meijers. J.H.C .. Bootsma. D .• and Gcurts van Kessel, A. (1990) The cellubr retinoic-acid-binding protein is expressed in tissues associated with retinoic-acid-induced malformations. Development 110:371-378.

van Straaten. H.W.M .. Thors. E. Hoessels. E.L .. Hekking. J.W.M .• and Drukker. J. (1985) Effect of.:i notochordal implant on the early morphogenesis of the neuml tube and neuroblasts. Dev. 8iol. 110:247-254.

van Str:l:lten. H.W.M .. Hekking. J.W.M .. Wiertz-Hoessels. E.J,L,M" Thors. F .. and Drukker. J. (1989) Effect of the notochord on the differentiation of a floorplate area in the neural tube of the chick embryo. Anat. Embryo\. 177:317-324.

Verwoerd. C.D.A .. and van Oostrom, c.G. (1979) Cephalic neuml crest and pbcode..~. Adv. Anal. Embryo!. Cell BioI. 58:1*75.

Vincent, M .. Duband. J .. and Thiery. J. (1983) A cell surface determinant expressed early on migrating avian neural crest cells. Dev. Brain Res. 9:235-238.

Vincent, M .. and Thiery. J. (1984) A cell surface marker for neural crest and plaeodal cell~: further evolution in peripheral and central nervous system. Dev. BioI. 103:468-481.

Vogel. D.L.. and Model. P.G. (1977) Development of the sympathetic system in the Mexican axolotl. Ambystoma mexicanum. Dev. BioI. 56:76-96.

Webster. W.S .. Johnston. M.C .. Lammer. EJ .. and Sulik. K.K. (1986) Isotretinoin embryopathy and the emnial neural crest: An in vivo and in vitro study. Dev. BioI. 6:211-222.

Weston, J.A. (1963) A ro.dioautographic analysis of the migration and Iocali7..ation of trunk neural crest cells in the chick. Dev. BioI. 6:279~3l0.

Weston. J.A .. and Butler. S,L. (1966) Temporal factors affecting localization of neural crest cells in the chicken embryo. Dev. BioI.

Wilkinson. D .. Bhatt. S .. Cook. M .. Boncinelli. E .. and Krumlauf. R. (1989) Segmental expression of Hox 2 homeobox-containing genes in the developing mouse hindbrain. Nature 34:405-409.

Wilson. H.C .. and Milos. N.C. (1987) The effects of various nutritional supplements on the growth. migration -'1nd differentiation of Xenopus laevis neural crest cells in vitro. In Vitro Cell Dev. BioI. 23:323-331.

Wolgemuth, OJ .. Behringer. R.R .. Mostoller. M.P .. Brinster. R.L .. and Palmiter. R.D. (1989) Transgenic mice overexpressing the mouse homeobox-eonmining gene Hox*1,4 exhibit abnormal gut development. Nature 337:464467.

39

Yntema. C.L .. and Hammond. W.S. (1954) The origin of intrinsic ganglia of trunk viscera from vagal ncuml erest in the chick embryo. J. Comp, Neurol. 101:515"541.

Ziller. c.. Fauquct. M .. Kalchcim. C .. Smith. J" and LcDouarin. N.M, (1987) Cdl lincage in peripheral nervous system ontogeny: medium"induced modulation of neuronal phenotypic expression in neural crcst cell cultures. Dcv, Biol. 120:101"111.

40

Chapter 2

Introduction to the enteric nervous system

2.1. Evolutionary aspects of intestinal motility

The development of specialized food-transport systems preceded the origin of vertebrates and

the vertebrate neural crest. Early prevertebrates were small, marine, filter-feeding organisms,

which had no specialized food.transport mechanisms and depended solely on diffusion for

their nutrition. But with the development of a muscular pharynx in protochordates. the first

primitive food-transport system became available. The functioning of such a system can be

illustrated by the pharyngeal apparatus in the nematode Caenorhabditis elegans. The pharynx

of C. elegans, which feeds on bacteria. consists of three functional parts (Fig. 7) (Avery and

Horvitz. 1989). The first part is formed by the corpus, which consists of two types of large

muscles, i.e. the procorpus (M3) and the metacorpus (M4). The second part consists of one

Figure 7: Anatomy and function of the pharynx of C. Elegans. A) The pharynx is divided into three functional parts, the corpus, the isthmus, and the tenninal bulb. The corpus is subdivided into the procorpus and metacarpus. There are five types of large muscles in the pharynx, arranged from anterior to posterior: m3 in the procorpus, m4 in the metacorpus, m5 in the isthmus, and m6 and m7 in the terminal bulb. B) Pumping: a pump consists of a nearly simultaneous contraction of the corpus, anterior isthmus, and terminal bulb, followed by relaxation. Corpus and isthmus muscles are radially orientated, so the lumen opens when they contract, sucking in liquid and suspended bacteria. Terminal bulb muscle contraction inverts the grinder, breaking bacteria that are in front of the grinder, and passing debris back to the intestine. Relaxation returns the grinder to its relaxed position and allows the lumen of the corpus to close, expelling liquid. Bacteria are trapped in a filter in the back of the corpus. C) Isthmus peristalsis: the feeding cycle is closed by a peristaltic contraction of the posterior isthmus muscles, which carries bacteria from the back of the corpus to the grinder. Anterior is to the left. (From L. A very and H.R. Horvitz, 1989, Neuron 3:473-485)

lumen grinder

t-m,.l-E:):i: :::::::::;m':=fjj, corpus

proeorpus

to

\.....E (' , ~f (' ,

1

m"'. corpus

isthmus

:3

®

3 pumping

, , [ terminal:

I bulb 'I

=3 3)

(3

G isthmus peristalsis

43

large muscle called the isthmus (M5). The third part consists of two muscles (M6 and M7)

and is called the terminal bulb. Feeding is accomplished by two separately controlled muscle

motions: pumping and isthmus peristalsis. During a pump all muscles of the pharynx contract

nearly simultaneously followed by relaxation. Because of the radial orientation of M3-5. this

contraction leads to the opening of the lumen of the corpus and anterior isthmus, whereas the

posterior isthmus remains closed (Albertson and Thomson, 1976). This results in the sucking

in of liquid and suspended bacteria, which accumulate in the back of the corpus (Seymour et

al., 1983). During the same pump the terminal bulb contracts, resulting in the crushing of

bacteria that are in front of the grinder, a specialized structure in the terminal bulb. The debris

is subsequently passed back to the corpus (Doncaster. 1962). Bacteria were transported from

the corpus to the grinder during a previous feeding cycle which ended with a peristaltic

contraction of the posterior isthmus muscles (isthmus peristalsis). All these coordinated

pharyngeal muscle contractions are regulated by the pharyngeal nervous system consisting of

14 types of neurons. adding up to a total of 20 neurons (Albertson and Thomson, 1976). It

has been found, however, that the muscles involved in pumping can function autonomously

without any pharyngeal neurons being present, whereas for isthmus peristalsis the presence

of only one specific neuron is needed (Avery and Horvitz. 1989). Absence of this neuron

leads to an accumulation of bacteria in the corpus, resulting in stuffed larvae (a phenotype

somewhat resembling megacolon in vertebrates).

These data indicate that the main role of the pharyngeal nervous system in C. elegans

is merely to regulate the frequency of pumps and the precise timing of muscle contractions

in response to environmental and physiological cues. without being necessary for pharyngeal

function itself. Therefore this pharyngeal nervous system should be considered more of a

eNS-related extrinsic nervous system than a neural crest-related intrinsic nervous system.

A far more elaborate food-transport system. consisting of a fore-. mid- and hindgut.

is found in insects. The fore- and hindgut are derived from ectodermal invaginations that

occur soon after gastrulation. whereas the midgut may receive contributions from different

embryonic germ layers including ectoderm. mesoderm and endoderm (Campos-Ortega and

Hartenstein. 1985). The development of the ENS in insects was studied most thoroughly in

the moth (Manduca sexta). The ENS consists of two small ganglia (the frontal and the

hypocerebral ganglion) lying dorsally on the foregut and an enteric plexus at the foregut­

midgut boundary, which are connected via the recurrent nerve (Copenhaver and Taghert,

1989). From these ganglia a nerve network arises that runs superficially along the rest of the

alimentary tract (Fig. 8). The two ganglia of the foregut and the recurrent nerve are derived

from three neurogenic zones in the foregut epithelium which differentiate shortly after foregut

formation and give rise to chains of cells that emerge onto the foregut surface (Copenhaver

and Taghert, 1991). As these cells emerge from the epithelium. they briefly become

mitotically active, dividing once or twice, and then gradually coalesce into the ganglia and

44

Figure 8: Cellular domains within the ENS of Manduca sexta. Two distinct populations of enteric neurons can be distinguished by their position and organization. One group of about 70 neurons forms a pair of small ganglia on the anterior surface of the foregut: the frontal ganglion (FG), which is connected to the brain lobes (BR) via the frontal ganglion connections (FGC), and the hypocerebral ganglion (HG), which is continuous with the recurrent nerve (RN) that lies mid-dorsally on the foregut surface. Paired nerves also connect the recurrent nerve to the neurohemal organs of the brain, the corpora cardiaca­corpora allota (CC-CA). A scond group of about 300 neurons (the EP cells) occupies the enteric plexus, a branching set of nerves that extend along eight longitudinal muscle bands on the midgut (only the dorsal muscle bands are shown: £l-L2 and R1-R2). (From P.F. Copeniw.ver, 1993, Development 117:59-74) Enteric

Plexus

nerves of the anterior ENS. Enteric glial cells are also generated from these neurogenic zones

as a distinct population of precursor cells (Copenhaver, 1993). These cells, which are among

the last to emerge from zones 2 and 3, remain mitotically active for a prolonged period of

time. At the end of this phase of neurogenesis, an ectodennal placode invaginates from the

foregut epithelium, at the site of the third neurogenic zone (Copenhaver and Taghert, 1990;

Copenhaver and Taghert. 1991). This ectodermal placode fonns a packet of enteric plexus

cells in the form of a triangular cluster at the foregut-midgut boundary with its anterior tip

connected to the recurrent nerve. Subsequently, this cluster of cells is transformed into a

45

mature enteric plexus by a sequence of migratory events. During the first slow phase of

migration, which is circumferentially directed, the triangular cluster spreads bilaterally as a

cell sheath around both sides of the foregut. In the subsequent fast phase of migration the

packet becomes disrupted by the abrupt dispersal of small subgroups of presumptive neurons

that stream out along an array of pathways on both the fore- and midgut to innervate the

visceral musculature (Copenhaver and Taghert, 1989). The dispersal of glial cells proceeds

along these same pathways (Copenhaver, 1993). The pattern of cell migration observed during

the formation of the ENS in the moth shares several important features with the migratory

behavior of cells derived from the vertebrate neural crest: migration occurs in multiple phases

(Tosney, 1978; Thiery et al., 1985; Newgreen and Erickson, 1986) and cells do not

necessarily follow particular pathways, but choose their pathway by reference to local cues

encountered in the course of their dispersal (Tosney, 1978; LeDouarin et al., 1984; Gershon,

1987; Copenhaver and Taghert, 1989). This shows that formation of the ENS in insects,

which lack the vertebrate neural crest, involves a developmental strategy which is clearly

distinct from neurogenesis in the insect eNS and closely resembles the generation of enteric

neurons in vertebrates.

2.2. Structure and ultrastructure of the enteric nervons system

The vertebrate alimentary tract contains a number of specialized regions each involved in a

particular function (Fig. 9). Although the general build up of the various regions of the

alimentary tract are the same, there are specific differences between the various vertebrate

species related to dietary variation (herbivore. carnivore Or omnivore). The morphology of the

gut wall varies between the different regions of the gut within one species and within one

region of different species, but in general consists of the luminal epithelium, mucosa.

submucosa, circular smooth muscle layer, longitudinal smooth muscle layer and serosal

epithelium (Fig. 10) (Furness and Costa, 1980).

The ENS of vertebrates is made up of an extrinsic and an intrinsic component. The

extrinsic component consists of a parasympathetic and a sympathetic division. The

parasympathetic innervation derives from the vagal nerve and is capable of enhancing

peristaltic activity. The sympathetic innervation derives from the splanchnic nerves and

terminates in the large abdominal autonomic plexuses, such as the celiac plexus innervating

the stomach and small intestine. This division of the extrinsic nervous system has a general

suppressive action on peristaltic activity.

The intrinsic component of the ENS consists of nerve plexuses and their

interconnecting fibers, embedded in the wall of the gut. The two principal plexuses are the

myenteric (Auerbach) and submucous (Meissner) plexuses. The myenteric plexus is confined

46

Figure 9: Schematic drawing of the chicken digestive tract, indicating its various pans. Oe= esophagus; VS= proventriculus; Ge= gizzard; Du= duodenum; Bi= bile duct; 1= jejunum; Om= umbilicus; 1= ileum; Ca= ceca; R= rectum.

Pa.N.

Ca

Pe.N.

Me.

no. )\vs 4J~

J

Om

R

Figure 10: Diagram of the arrangement of the enteric nervous system in a transverse section of the bowel. Me.=mesentery; M.P.=mucosal plexus; Pa.N.=paravascular nerve; Pe.N.=perivascular nerve; S.N.=subserous nerve; S.P.=submucous plexus. (From Furness and Costa. 1980)

47

to the space between the circular and longitudinal muscle layers and, because of this position,

undergoes enormous variations in size and shape during intestinal contraction. Its ganglia

generally contain a large number of nerve cell bodies, which are contained in a meshwork that

is quite characteristic and can be readily identified in anyone area from a particular species

on account of the thickness of the main nerve bundles and their branching pattern. The

submucous plexus lies embedded in the connective tissue between the circular smooth muscle

layer and the mucosa. Compared to the myenteric plexus, ganglia are smaller, containing

fewer neurons, whereas the meshes are larger and more irregular. Sometimes two plexuses

can be recognized within the submucosa called the plexus of Henle and the plexus of

Meissner, which can be physically separated. The plexus of Henle appears to be a motor

plexus as it contains neurons which closely resemble those found in the myenteric plexus. The

plexus of Meissner is composed of small neurons resembling the neurons of the cerebro-spinal

ganglia and which therefore appear to be sensory. In both the myenteric and the submucous

plexuses, there are considerable variations in the size and shape of the ganglia and meshes,

not only between different species, but also between different regions of the gut within one

species, and between the mesenteric and the anti-mesenteric side within one region (Weyns,

1988).

The ultrastructure of the ENS differs considerably from that oC other autonomic

ganglia and in many respects resembles more closely the CNS. The ENS contains a large

number of neurons and a remarkable diversity of neuronal cell types. The perikarya are

irregularly shaped and characterized by large eccentrically placed nuclei with fine granular

nucleoplasm, prominent nucleoli and sparse condensations of chromatin. The ENS also

contains numerous glial cells which closely resemble eNS astrocytes, not Schwann cells

(Gabella. 1971; Cook and Bumstock. 1976b). containing more glial fibrillary acidic protein

(Bjorklund et al .. 1984; Rothman et al .. 1986) and extending more processes (Erde et al .•

1985). It has been suggested that the enteric glial cells may be important in conferring

structural stability to the ganglia, holding them together and at the same time allowing

structural rearrangement of the ganglia during muscle contraction. In contrast to axons of

peripheral nerves, enteric axons are not individually enveloped by a glial process, but bundles

of enteric axons are invested by glial processes so that neurites abut on one another. Within

the enteric ganglia, perikarya and glial cells are tightly packed with a virtual absence of an

intraganglionic extracellular space, whereas other autonomic ganglia are loosely organized

containing collagen, blood vessels. fibroblasts. macrophages and mast cells in addition to

neurons and supporting cells. Finally. enteric ganglia are isolated from the surrounding tissue

by a basal lamina which develops after complete formation of the ganglia. This basal lamina

may function as a blood-plexus boundary. analogous to the blood-brain barrier. As already

stated, the ENS harbors a large variety of neurons. which have been classified

48

Table 3: Putative and established neurotransmitters in the ENS

Class

Acetylcholine

Noradrenaline

Biogene amines

Amino acids

Purines

Neuropeptides

Subclass

serotonin

norepinephrine

GABA (gamma-aminobutyric

acid)

ATP

Substance P

VIP (vasoactive intestinal

polypeptide)

peptide histidine·isoleucine

somatostatin

calcitonin gene-related peptide

neuropeptide Y

pancreatic polypeptide

enkephalins

beta-endorphin

dynorphin

adreno corrico traphin

alpha-melano trophin

galanin

cholecystokin

gastrin-releasing peptide

bombesin

arginine vasopressin

Reference

(Costa et al., 1982; Furness and

Costa, 1982)

(Bumstock and Costa, 1975)

(Jessen et aL. 1986)

(Burnstock, 1972)

(Costa et al., 1981)

(Furness et al., 1981; Costa and

Furness. 1983)

(Bishop et al.. 1984)

(Costa et aI .• 1980; Furness et aI .. 1980)

(Rodrigo et aJ.. 1985)

(Furness et al .• 1983.1; Daniel

et aI .• 1985)

(Furness et .11.. 1983a)

(Furness et a1.. 1983b; Costa et

aI .• 1985)

(Wolter. 1985a)

(Dalsgaard et aI., 1983: Costa

et aI .• 1985)

(Wolter. 1985b)

(Woiter, 1985b)

(Ekblad et aI., 1985: Melander

et aL. 1985)

(Hutchison et aI.. 1981)

(Hutchison et a1.. 1981: Costa

et a1.. 1984)

(Hanley et .11., 1984)

49

morphologically (Dogiel, 1895; Gunn, 1959; Gunn, 1968; Baumgarten et al" 1970; Feher and

Vajda, 1972; Cook and Bumstock, 1976a), These different neurons contain an abundance of

putative and established neurotransmitters, not seen in any other region of the autonomic

nervous system and most of which are also active in the brain (Table 3) (Furness and Costa.

1980; Gershon, 1981; Gershon and Erde, 1981), Besides the two classical neurotransmitters,

acetylcholine and noradrenaline, there is a large and complex peptidergic system distributed

throughout the lengths of the gut. More recent evidence for the coexistence of different

substances in various combinations within the same neuron. have made the picture of enteric

innervation even more complex CMakhlouf, 1985). The presence of a large variety of neuronal cell types with a characteristic distribution pattern within the gut, and with a well-defmed set

of actions, suggests that the ENS could function as a 'mini-brain', under the general influence

of the eNS, but able to function by itself.

2.3. Development of the enteric nervous system in amphibians

Development of the ENS in amphibians has been studied most extensively in Xenopus by

constructing interspecific chimeras between Xenopus laevis and Xenopus borealis embryos

(Sadaghiani and Thiebaud, 1987; Epperlein et al" 1990), In this way. it has been established

that the enteric neurons originate from the neural crest adjacent to the posterior

rhombencephalon and the anterior spinal cord (somites 1 and 2), These neural crest cells

migrate along a ventromedial pathway between the somites and the neural tube on their way

to the primitive gut (Epperlein et al" 1990). Sadaghiani and Thiebaud (1987) showed that

these cells are first found scattered in a region between the last branchial arch and the

posterior part of the pronephros, and subsequently migrate along dorsal root fibers and

through the body of the vagus ganglion complex. They continue migration from the lower part

of this ganglion complex by means of the recurrent-intestinal nerve, which runs over the

glottis musculature, and penetrates the wall of the esophagus and larynx. Then they

translocate craniocaudally through the splanchnopleural layer of the intestinal tract and its

accessory glands and differentiate into mature enteric neurons.

2.4. Development of the enteric nervous system in birds

The avian gut is formed by lateral infolding of the endoderm and the splanchnic mesodermal

epithelium. This results in the formation of a tube, consisting of a lumen surrounded by

endodermal and splanchnic epithelia. The mesenchymal cells of the gut are formed through

de-epithelialization of cells from the splanchnic epithelium, thus forming a loose mesenchyme

50

beween the two epithelia which later gives rise to the muscles and connective tissue of the

gut.

The origin of the enteric neurons has long been an issue of debate (see (Andrew,

1971) for review), but it is now generally agreed upon that the enteric neurons are mainly

derived from the vagal neural crest (Yntema and Hammond. 1954; LeDouarin and Teillet,

1973; Allan and Newgreen, 1980). In their fate·map study, LeDouarin and Teillet (1973) also

found evidence for a contribution of the lumbosacral neural crest, caudal to somite 28. to the

ENS in the postumbilical gut. This sacral contribution to the postumbilical ENS has since then

been confirmed in a number of studies (Pomeranz and Gershon. 1990~ Pomeranz et al., 1991a~

Serbedzija et aI., 1991), but its precise fate has not yet been established. It has been shown

that in the absence of vagal neural crest cells, sacral neural crest cells are not capable of

giving rise to enteric ganglia (Yntema and Hammond, 1954) (Peters-van der Sanden et al.,

unpublished). The sacral neural crest does give rise to the ganglion of Remak (Yntema and

Hammond, 1953; Yntema and Hammond. 1955; Teillet, 1978). a ganglionated nerve which

is only present in birds and which belongs to the autonomous nervous system. It runs parallel

to the gut in the mesentery and extends from the duodenal-jejunal junction, where it is

connected with the celiac plexus, to the cloaca, where it joins the pelvic plexus (Nolf, 1934).

Although the previous paragraph indicates that the vagal neural crest is the principal

source for enteric neurons in vivo, there is some debate whether other axial levels of the

neural crest may also be capable of giving rise to enteric ganglia in an experimental system.

In a number of studies, aneural hindgut, not containing neural crest cells, was cocultured with

neural crest segments derived from various axial levels, on the chorioallantoic membrane of

a host embryo (LeDouarin and Teillet, 1974; Smith et aI., 1977; Teillet et aI., 1978; Newgreen

et aI., 1980; Smith-Thomas et aI., 1986). These studies described the presence of enteric

neurons in the gut, but the number of neurons foun~. highly depended on the experimental

conditions used. Furthermore, in all of these studies melanocytes were found in the gut which

were mainly present at the sites were ganglion formation would normally occur. In heterotopic

quail-chick chimeras, in which the chicken vagal neural crest was replaced with quail neural

crest from the adrenomedullary level (SI8-24), quail cells were found along the entire

digestive tract (LeDouarin and Teillet, 1974). But whereas in the preumbilical gut these quail

cells had differentiated into enteric neurons, in the postumbilical gut these quail cells

exclusively differentiated into melanocytes. In the reciprocal experiment, in which quail vagal

neural crest was transplanted to the adrenomedullary level of a chicken embryo, they not only

found quail cells in the normal adrenomedullary derivatives, but also in the postumbilical gut,

providing further evidence for a specific interaction between vagal neural crest cells and the

postumbilical gut regarding ENS formation. These experiments clearly indicate that. although

neural crest cells from various axial levels can migrate through the gut and home to the

correct sites, differentiation into enteric neurons, at least in the postumbilical gut, is confined

51

to vagal neural crest cells.

The vagal neural crest fonns a transitional zone between the cranial and trunk neural

crest. Cells from this region mainly emigrate from the neural tube between stages 9 and 12

and migrate along two principal pathways. Cells from the anterior vagal neural Crest from the

level of the mid-otic vesicle down to the third somite predominantly migrate along a

dorsolateral pathway on their way to the three most caudal pharyngeal arches (Kuratani and

Kirby. 1991; Miyagawa-Tomita et al .. 1991). Ectomesenchymal derivatives of the vagal neural

crest, such as the cardiac outflow tract and the mesenchymal components of the thymus and

parathyroids, are exclusively derived from this part of the vagal neural crest (Kirby et al .•

1983; Bockman and Kirby. 1984). Cells from the posterior vagal neural crest caudal to somite

2, migrate along a ventrolateral pathway through the rostral part of the somites (Bronner­

Fraser et al., 1991; Miyagawa-Tomita et al., 1991). It has not been established along which

migration pathway neural crest cells reach the gut. There is some indirect evidence that vagal

neural crest cells migrate to the gut via the most caudal pharyngeal arches (Ciment and

Weston, 1983: Payette et al .. 1984: Tucker et al., 1986), but migration along a ventrolateral

pathway has also been suggested (LeDouarin, 1982: Noden. 1988; Serbedzija et al., 1991).

It is generally agreed upon that neural crest cells enter the foregut at E2.5 (stage 16

or l7 HH) (LeDouarin and Teillel. 1973: Tucker et al., 1986), but the exact level of entry is

hitherto unknown. These cells subsequently translocate in a craniocaudal direction, either

through active migration or through passive displacement due to bowel elongation, following

an exact time schedule (LeDouarin and Teillet, 1973; Tucker et al., 1986). There are few

studies on the migration of neural crest cells through the gut. Tucker et al. (1986) described

that neural crest cells migrate superficially, using the splanchnic epithel~um as a substrate,

whereas LeDouarin (1982) found that neural crest cells migrate as dispersed cells through the

loose mesenchyme of the gut. Epstein et al. (1991) observed a complex network of neural

crest-derived cells in the gut. indicating that neural crest cells interact with each other during

migration. These studies. however, all agree upon the importance of the enteric mesenchyme

in· the guidance of neural crest cell migration, in the homing at specific sites and in the

differentiation into the various types of enteric neurons. A number of molecules within the

enteric mesenchyme have been described, which might playa role in one or more of these

processes, i.e. fibronectin (Tucker et al .• 1986), larninin (Pomeranz et al .. 1991b), and HNK-1-

carrying glycoproteins (Luider et al.. 1992). but their relative importance has yet to be

established.

2.5. Development of the enteric nervous system in mammals

Development of the mammalian ENS has been studied in human (Okamoto and Deda. 1967),

52

and, more extensively, mouse embryos. Although ENS formation in mouse embryos is in

principal the same as described for avian embryos, some apparent discrepancies should be

mentioned here. Whereas in chicken embryos neural crest cells first enter the foregut and

subsequently colonize the rest of the gut in a craniocaudal order, studies in mouse embryos

suggested simultaneous colonization of fore- and hindgut at E9 (Rothman and Gershon, 1982;

Gershon et al., 1984; Rothman et al., 1984: Rothman et al., 1986). Other studies. however,

are in disagreement with these results and find a craniocaudally directed colonization process

in mice as well (Nishijima et al .. 1990; Kapur et al., 1992). A second difference lies in the

fact that in mouse embryos transient catecholaminergic neurons are present in the wall of the

gut, which are lost upon the arrival of sympathetic nerves (Rothman et al., 1987). In the avian

gut, these transient catecholaminergic cells could only be observed in vitro, after dissociation

of the gut and culturing of the enteric precursors outside the enteric microenvironment. These

cells, however, were never found in vivo.

Much of the knowledge of mammalian ENS development comes from the study of

animal models in which ENS formation is disturbed. Spontaneous occurrence of aganglionosis

has been reported in mice, rats and horses. In mice. three different genetic mutations have

been described. which all result in congenital megacolon combined with pigment

abnormalities. The dominant spotting mutation is autosomal dominant. Heterozygotes are

characterized by White-spotting and a deficiency of enteric neurons in the colon, whereas

homozygotes die prior to gestational day 13 (Lane and Liu, 1984). Mutations in the autosomal

recessive piebald-lethal (ST) gene lead to a defect in the migration of cells from the neural

crest (Webster, 1973). Sl/Sl homozygotes are white-coated except for patches of black

pigment and invariably develop megacolon early in life (Lane, 1966). Animals often die from

diarrhoea and enterocolitis before breeding age is attained, making the establishment of

colonies very difficult. The lethal-spotted (Is) strain resembles the piebald-lethal, but

homozygotes tend to survive longer making them more suitable for study (Lane, 1966). This

gene is also transmitted in an autosomal recessive way.

In lethal-spotted mice, the distal 2 rum of the gut is aganglionic. Although this

aganglionic segment lacks intrinsic innervation, it does receive many noradrenergic nerve

fibers derived from neurons, whose cell bodies lie outside the gut (Payette et al., 1987). These

extrinsic nerve fibers are characteristic of non-enteric peripheral nerves, because their axons

are enveloped individually by Schwann cell processes with basal lamina and collagen­

containing endoneurial sheaths (Tennyson et al .. 1986c). The lack of intrinsic innervation in

these mutant mice could be due to a defect in the neural crest or in the target organ. The

occurrence of pigment abnormalities combined with ENS defects could point to a neural crest

defect. In coculture experiments, however, neural crest cells from the foregut of lslls mice

were found to be able to colonize the hindgut of both avians and control mice, whereas

neither avian nor control mice crest was able to colonize lslls hindgut (Jacobs-Cohen et al.,

53

1987). Additional evidence that the primary defect in Is/ls mice is not neuroblast autonomous.

comes from a study of Kapur et al., who made aggregation chimeras of cells derived from

ls/ls mice and mice carrying the DBH-IacZ transgene (a marker for enteric neurons) (Kapur

et al., 1993). They found that when more than 20% of the cells in chimeric mice were wild­

type, the Islls phenotype was rescued. In addition, these rescued mice had mixtures of both

Islls and wild-type neurons throughout the gut, including the distal rectum.

The hindgut of lslls mice shows an extraordinary overabundance and maldistribution of laminin and collagen type IV (Tennyson et al., 1986a; Tennyson et al., 1986b). These

extracellular matrix molecules. which are normally found in basal lamina beneath the mucosal and serosal epithelia. and around smooth muscle cells and ganglia (Pomeranz et al., 1991b),

are now present in a broad zone encompassing the entire mesenchyme of the aganglionic

segment of the gut. This has led to the hypothesis that these basal lamina components may

normally act as stop signals for neural crest cells, which carry specific receptors for these

molecules (Pomeranz et al., 1991b), inducing them to cease migration, extend neurites and

withdraw from the cell-cycle. Overabundance of these basal lamina components could then

be interpreted as an extension of the normal mechanism in which neural crest cell migration

is stopped prematurely. while not inhibiting axon growth.

Comparison of various animal species shows that the basic principles of ENS

development are very similar. Each of these species, however, provides us with different tools

to study the various aspects of ENS development.

2_6. Clinical disorders of the enteric nervous system

Study of spontaneous mutations might provide inside into normal embryonic deVelopment.

An important congenital malformation involving the ENS in man, is congenital intestinal

aganglionosis or Hirschsprung disease (HSCR). It is characterized by the absence of enteric

ganglia in the most distal part of the bowel, combined with the presence of hypertrophic

extrinsic nerve fibres. It was first described by Dr. Harald Hirschsprung (1830-1916), who

described two newborns that presented with characteristic clinical features: severe defecation

problems from birth onwards. increasing abdominal distension, enterocolitis with ulcers, a

gradually declining general condition leading to death at the age of 7 and 11 month

respectively. At first, the defect was thought to lie in the dilated and hypertrophied part of the

colon. but Swenson was among the first to recognize that the primary defect lay in the distal

non-dilatated colon (Swenson et al., 1949). In 1948, he introduced a new surgical technique,

and since then children born with HSCR can be successfully operated, although some patients

remain obstipated after surgery.

HSCR can be classified based on the length of the aganglionic segment, which always

54

encompasses the internal anal sphincter and sometimes extends well into the small bowel.

Short-segment or classical HSCR is the best described congenital ENS malformation. The

aganglionosis begins with the internal anal sphincter and extends proximally including the

rectum and part of the sigmoid colon. In long-segment HSCR. aganglionosis extends beyond

the sigmoid colon, sometimes involving the entire colon (total colonic aganglionosis or

Zuelzer-Wilson disease), or even part of the small bowel. The longest aganglionic segment

is found in total intestinal aganglionosis, a rare congenital malformation in which enteric

neurons are lacking from the duodenum downwards to the anus. HSCR has an estimated

overall population incidence of 1:5000 live-born children. a sex ratio of 3.9 to I. males to

females, and an overall risk to siblings of 4%.

Genetic study of HSCR has proven difficult. because of the very limited availability

of large pedigrees. Until fairly recently, patients with HSCR hardly gave rise to offspring.

Furthermore. having a child with HSCR often results in curtailment of child bearing. A

positive family history has been described in approximately 7% of all cases. The high number

of sporadic cases and the fact that the disease is four times more frequent in boys than in girls

suggested a sex-modified multifactorial mode of inheritance. involving multiple genes

(Passarge. 1983). Badner et al. (1990) performed complex segregation analysis on 487

probands and their families and showed that for short-segment HSCR the inheritance pattern

was equally likely to be either multifactorial or due to a recessive gene with low penetrance.

For long-segment HSCR. however, they found that the mode of inheritance was most

compatible with an autosomal dominant gene with incomplete penetrance.

Deletions of the long arm of chromosome 10 (IOq 11) have been found in a patient

with HSCR (Martucciello et al .• 1992). Recently a gene for LS-HSCR was mapped to the

proximal long arm of chromosome IO (lOq 11.2) (Angrist et al .• 1993: Lyonnet et al .• 1993).

Edery et al. (submitted for publication) described the c-RET proto-oncogene as the closest

genetic marker with respect to the disease locus. suggesting that this proto-oncogene might

be a candidate gene for HSCR. c-RET has also been shown to account for multiple endocrine

neoplasia type 2A (MEN 2A). characterized by the occurrence of multiple neural crest-related

tumors, such as medullary thyroid carcinomas and pheochromocytomas (Donis-Keller et al..

1993; Mulligan et aI .. 1993). The finding that transgenic mice carrying a null mutation of the

ret-l gene have total intestinal aganglionosis, further suggests that the RET-J gene might be

involved in the pathogenesis of long-segment HSCR (Dr. V. Pachnis. personal

communication).

In mice, the ret~l gene maps to chromosome 14 where the piebald lethal gene has

been localized (Badner et al.. 1990). The existence of three different mouse models. however.

each with a single gene defect, could point to the existence of more than one gene locus for

HSCR (Lane, 1966). Identification of other possible candidate genes in humans comes from

the study of chromosomal abnormalities associated with HSCR. Particularly striking is the

55

high incidence of HSCR with trisomy of chromosome 21, known as Down syndrome

(Passarge, 1967). HSCR was also found to be associated with various deletions on the long

arm of chromosome 13 (Sparkes et al .• 1984; Lamont et al., 1989; Bottani et al., 1991). In

these patients an association with, craniofacial abnormalities and mental retardation has been

described.

The various forms of HSCR are sometimes associated with other anomalies. For the

long-segment ENS disorders associated anomalies have rarely been described, apart from

occasional kidney abnormalities in total intestinal aganglionosis (DiLorenzo et al., 1985), and

neuroblastomas in total colonic aganglionosis (Michna et al., 1988). For classical

Hirschsprung disease the reported incidence of associated anomalies varies from 2.5 to 29.8%

depending on the diligence with which they are sought and the manner in which they are

reported (Graivier and Sieber. 1966; Swenson et al., 1973; Klein et al., 1984; Spouge and

Baird, 1985; Ikeda and Goto, 1986; Ryan et al., 1992). Many of the associated anomalies are

related to the third and fourth pharyngeal arches, receiving a contribution from the posterior

rhombencephalic neural crest. These anomalies mainly entail heart defects, often characterized

as tetralogy of Fallot (Lammer and Opitz, 1986; Larsen, 1993), whereas abnormalities of the

thymus, which also rec,eives a contribution from the posterior rhombencephalic neural crest.

have never been reported (van Dommelen et al., chapter 3.7).

Next to the described forms of congenital aganglionosis, there are a number of other

congenital disorders of the ENS worth mentioning here. Intestinal hyperganglionosis. or

. neuronal intestinal dysplasia', is characterized by an above-average number of neurons in the

enteric ganglia and the presence of hyperplastic parasympathetic nerve trunks (Meier-Ruge,

1971; Howard and Garret, 1984). Colonic hypoganglionosis is characterized by a below­

average number of enteric neurons and nerve fibers in the myenteric plexus (Meier-Ruge,

1971). Hypoganglionosis is described mostly in the transition zone between normal and

aganglionic bowel of patients with HSCR (Gherardi, 1960; Walker et al., 1966; Garret et al.,

1969; Meier-Ruge, 1969). There are a few reports of zonal aganglionosis (Tiffin et al .. 1940;

Haney et al., 1982; Tagushi et aI., 1983; Seldenrijk et al., 1986), characterized by the presence

of ganglionic segments within the aganglionic bowel. but its existence has been questioned

(Yunis et al., 1983). Intestinal pseudo-obstruction is characterized by a functional obstruction

of the ganglionic bowel and has often been associated with neuronal immaturity (Bughaighis

and Emery, 1971; Erdohazi, 1974; Tanner et al., 1976). Besides congenital ENS disorders,

there are also a number of acquired disorders of the ENS. In Chagas' disease enteric ganglia

are lost secondary to a protozoan infection (Meneghelli, 1985). In patients with Parkinson's

disease, ENS functioning can also be affected resulting in acquired megacolon (Kupsh.")' et al.,

1987). The finding of cytoplasmic hyaline inclusions in myenteric neurons which are identical

to Lev.ry bodies found in the brain of Parkinson patients, further suggests a similarity between

CNS and ENS neurons.

56

2.7. Conclusions

Study of the evolutionary origin of the ENS showed that ENS development preceded the

origin of the vertebrate neural crest. In the moth, enteric neurons and glial cells arise from

ectodermal placodes, in a way that shares several important features with the migratory

behavior of vertebrate neural crest cells. The fact that no pharyngeal arch system has yet

developed in the moth, indicates that ENS precursors do not have to migrate through the

pharyngeal arches in order to be able to differentiate into enteric neurons.

Comparing various vertebrate species, we showed that, in general, basic principles of

ENS development are very similar. Each species, however, offers certain advantages to study

specific aspects of ENS development. Chicken embryos are well suited for experimental

embryological manipulations, and in this species specific markers are available both for neural

crest cells and for enteric neurons. Genetic study of intestinal aganglionosis can be performed

in mammalian embryos. In mice, three spontaneously occurring mutants led to the

identification of three recessive genes involved in ENS development. Recently described

markers for enteric neural crest cell precursors and enteric neurons in mice, could further

facilitate the study of the defects in ENS development occurring in each of these mutants. In humans, search for chromosomal abnormalities and linkage analysis in large pedigrees of

HSCR patients already led to the identification of the c-RET proto-oncogene as a candidate

gene for HSCR.

2.8. References

Albertson. D.O .. and Thomson, J.N. (1976) The pharynx of CaenorhcJ.bditis clegans. PhiL Trans. Roy. Soc. (Lond) B275:299-325.

Allan, D .. and Newgreen. D.F. (1980) The origln and differentiation of enteric neurons of the intestine of the fowl embryo. Am. J. Annt. 157:137-154.

Andrew, A. (1971) The origin of intramural ganglia. IV The origln of enteric ganglia: a critical rcview and discussion of the present state of the problem. J. Anat. 108:169-184.

Angrist. M .. Kauffman. E .. Slaugcnhaupt, A.. Matise. T.C .. Puffenberger. E.G .. Washington. S.S., Lipson. A .. Cass. D.T .. Reyna, T .. Weeks, D.E .. Sieber, W .• and Chakravarti. A. (1993) A gene for Hirschsprung disease (megacolon) in the pericentromeric region of human chromosome 10. Nature Genetics 4:351-356.

Avery. L .. and Horvitz. R. (1989) Pharyngcal pumping continucs after laser lcilling of the pharyngeal nervous system of C. elcgans. Neuron 3:473-485.

Sadner, JA, Sieber. W.K .. Garver, K.L. and Chakravarti, A. (1990) A genetic study of Hirschsprung disease. Am. J. Hum. Genet. 46:568-580.

Baumgarten. H.G .. Holstein. A.F .. and Owman. C. (1970) Auerbach's plexus of mammals and man: electron microscopic identification of three different types of neuronal processes in myenteric ganglia of the large intestine from rhesus monkeys. guinea pigs and man. Z. Zellforsch. Mikros. Anat. 106:376-397.

Bjorklund. H .. Dahl. D .. and Singer. A.M. (1984) Neurofilament and glial fibrillary acidic protein-related immunoreactivity in rodent enteric nervous system. Neurosci. 12:277-287.

57

Bockman, D.E .. and Kirby, M,L, (1984) Dependence of thymus development on derivatives of the neural crest. Science 223:498~500.

Bottaru, A" Xie, Y" Binkcrt, F .. and Schinzcl, A. (1991) A case of Hirschsprung discase with a chromosome 13 microdeletion, del(13){q32,3q33.2): potential mapping of one disease locus, Hum, Genet. 87;748·750.

Bronner~Fraser. M., Stem, C.D .. and Fraser, S. (1991) Analysis of neural crest ccll lineage and migration. J. Craniofac. Genet. Dev. BioI. 11:214-222.

Bughaighis, A.G" and Emery, J.L. (1971) Functional obstruction of the intestine due to neurological immaturity. Progt'. Pediatr. Surg. 3:37~52.

Campos-Ortega, lA., and Hartenstein, V. (1985) The embryonic development of Drosophila mclanogaster. Springer~ Verlag: Heidelberg.

Ciment, G .. and Weston, J.A. (1983) Enteric neurogenesis by neural crest~derivcd branchial arch mesenchymal cells. Nature 305:424-427,

Cook, R.D .. and Bumstock, G. (1976a) The ultrastructure of Auerbach's plexus in the guinea-pig. I. Neuronal clements. J. Neurocyto!. 5:

Cook, R.D .. and Burnstock. G. (1976b) The ultrastructure of Auerbach's plexus in the guinea-pig. II. Non-neuronal clements. J. Neurocyto!. 5:195-206,

Copenhaver. P.F .. and Taghert, P.H. (1989) Development of the enteric nervous system in the moth II. Stereotyped cell migration precedes the differentiation of embryonic neurons. Dev. BioL 131 ;85·101.

Copenhaver, P.F., and Taghert, P.H. (1990) Neurogenesis in the insect enteric nervous system: genemtion ofpre~migratory neurons from an epithelial placode. Development 109:17-28.

Copenhaver. P.F" and Taghert. P.R (1991) Origins of the insect enteric nervous system: differentiation of the enteric ganglia from a neurogenic epithelium. Development 113:1115-1132.

Copenhaver. P.E (1993) Origins. migration and differentiation of glial cells in the insect enteric nervous system from a discrete set of gli:ll precursors. Development 117:59~ 74.

DiLorenzo, M .. Yasbeck, S .. and Brochu. P. (1985) Agangiionosis of the entire bowel: four new cases and review of the literature. Br. J. Surg. 72:657~658.

Dogiel, A.S. (1895) Zur frage tiber die ganglion der danneflechte bei den saugetieren. Anat. Anz. 10:517-528.

Doncaster, C.c. (1962) Nematode feeding mechanisms. I. Observations on Rhr;.bditis and Pelodera. Nematologica 8:313~320.

Donis~Kelier. H .. Dou. S .. Chi, D .. Carlson, KM., Toshima, K" Lainnore. T.C .. Howe. J.R .. Moley. IF .. Goodfellow. P .. and Wells, S.A. (1993) Mutations in the RET proto-oncogene arc associated with MEN 2A and FMTC. Human Mo!. Gen. 2:851-856.

Epperlein. H.H .. Krotoski. D .. Halfter, W .. and Frey. A. (1990) Origin and distribution of enteric neurons in Xenopus. Anal. Embryol. 182:53~67.

Epstein. M,L .. Poulsen. KT .. and Thiboldeaux. R. (1991) Formation of ganglia in the gut of the chick embryo. J. Compo Neurol. 307:189~199.

Erde. S.M .. Sherman, D .. and Gershon. M.D. (1985) Morphology and serotonergie innervation of physiologically identified cells of the guinea pig's myenteric plexus. J. Neurosei. 5:617-633.

Erdohazi, M. (1974) Retarded development of the enteric nerve cells. Dev. Med. Child Neurol. 16:365·368,

Feher. E .. and Vajda, J. (1972) Cell types in the nerve plexus of the small intestine. Acta Morphol. Acad. Sci. Hung. 20: 13~25.

Furness, J.B .. and Costa, M. (1980) Types of nerves in the enteric nervous system. Neurosci. 5:1~20.

Gabella, G. (1971) Glial cells in the myenteric plexus. Z. Naturforseh. 26B:244-245.

Garret. JR. Howard. E.R" and Nixon. H.H. (1969) Histochemical diagnosis of Hirschsprung's disease. Lancet 11:436.

58

Gershon. M.D. (1981) The enterie nervous system. Ann. Rev. Neurosei. 4:227~27L

Gershon. M.D .. and Erde. S.M. (1981) The nervous system of the gut. Gastroenterology 85:929~937.

Gershon. M.D .. Payette. R .. Teitleman. G .. and Rothman. T.P. (1984) Neuronal commitment and phenotypic expression by developing enteric neurons. Wiley: New York. 181~204.

Gershon. M.D. (1987) Insights into neuronal development proVided by the bowel. Liviana Press: Padona. I08~144.

Gherardi. GJ. (1960) Pathology of the ganglionic~aganglionjc junction in congenital megacolon. Arch. PathoL 69:520~523.

Grnivier. L.. and Sieber. W.K. (1966) Hirsehsprung's disease and mongolism. Surgery 60:458-461.

Gunn. M. (1959) Cell types in the myenteric plcxus of the cat. J. Compo Neurol. 111:83"93.

Gunn. M. (1968) Histological and histochemical observations on the myenteric and submucosal plexuses of mammals. 1. Anat. 1 02:223~239.

Haney. PJ .. Hill. J.L .. and Sun. C.C. (1982) Zonal colonic agangJionosis. Pediatr. Radiol. 12:258-.261.

Howard. E.R .. and Garret. J.R. (1984) Hirschsprung's disease and other neuronal disorders of the hindgut. M.S. Tanner-and R.J. Stockss (ed). Intercept: New Castle upon Tyne. 12H37.

Ikeda. K .. and Goto. S. (1986) Additional anomalies in Hirschsprung's disease: an analysis based on the nationwide survey in Japan. Z. Kinderchir. 41:279"281.

Jacobs~Cohen. R,J .. Payette. RF .. Gershon. M.D .. and Rothman. T.P. (1987) Inability of neural crest cells to colonize the presumptive aganglionic bowel of Is/Is mutant mice: requirement for a permissive microenvironment J. Compo Neurol. 255:425-438.

Kapur. R.P .. Yost. C. and Palmiter. R.D. (1992) A transgenic model for studying development of the enteric nervous system in normal and aganglionic mice. Development : 16:167~175.

Kapur. RP .. Yost. C .. and Palmiter. R.D. (1993) Aggregation chimeras demonstr.:lte that the primary defect responsible for aganglionic megacolon in lethal spotted mice is not neuroblast autonomous. Development 117:993"999.

Kirby. M.L .. Gale. T.F .. and Stewart. D.E. (1983) Neural crest ccll~ contribute to normal aorticopulmonary septation. Science 220: I 059~ J 061.

Klein. M.D .. Coran. A.G .. Wesley. J.R .. and Drongowski. R.A. (1984) Hirschsprung's disease in the newborn, J. Pediatr. Surg. 19:370"374.

Kupsky. W,J .. Grimes. M.M .. Sweeting. J.. Bertsch. R .. and Cote. L.l. (1987) Parkinson's disease and megacolon: concentric hyaline inclusions (Lewy bodies) in enteric ganglion cells. Neurol. 37:1.253"1255.

Kuratani. S,C .. and Kirby. M.L. (1991) Initi~l migration ~nd distribution of the cardiac neural crest in the avian embryo: An introduction to the concept of the circumpharyngeal crest. Am. 1. Ana!. 191:215~227.

Lurnmer. E.J .. and Opitz. J.M. (1986) The DiGeorge anomaly as a developmental field defect. Am. J. Med. Gcnet. suppl. 2:1l3~127.

Lamont. M.A .. Fitchett. M .. and Dennis. N.R. (1989) Interstitial deletion of distal 13q associated with Hirschsprung's disease. J. Med. Genet. 26:100"104.

Lane. P,W. (1966) Association of megacolon with two recessive spotting genes in the mouse. J. Hercd. 57:181~183.

Lane. P,W" and Liu. H.M. (1984) Association of megacolon with a new dominant spotting gene (DOM): in the mouse. J. Hered. 75:435"439.

Larsen. W,J. (1993) Human embryology. W.R. Schmitt, M. Otway. and E. Bowman~schulman. (cds.), Churchill Livingstone: New York.

LcDouarin. N.M .. and Teillet. M. (1973) The migration of neural crest cells to the wall of the digestive tr~ct in avi~n embryo. J. Embryol. expo MorphoL 30:31-48.

59

LeDou::uin. N.M .. and Teillet. M. (1974) Experimental ano.1ysis of the migration and differentiation of neuroblasts of the autonomic nervous system and ofneuroectodennal mesenchymal derivatives, using a biological cell marking technique. Dev. BioI. 41:162-184.

LeDou::uin. N.M. (1982) The neural crest. C::unbridgc Univ. Press: Cambridge.

LeDouann. N.M .. Teillet. M.A. and Fontaine-PenL.". 1. (1984) Chimeras in the study of the peripheral nervous system of birds. N.M. Le Dou::uin and A. McLarens (eds). Academic Press: London.

Luider. T.M .. Peters-van der Sanden. M.lH .. Molenaar. lC .. Tibboel. D .. van der Kamp. AW.M .. and Meijers. C. (1992) Characterization of HNK-l antigens during the fonnation of the avian enteric nervous system. Development 115:561-572.

Lyonnet. S .. Bolino. A .. Pelet. A .. Abel. L.. Fekete. CN .. Briard. M.L .. Siu. V.M .• Kaariainen. H .. Martuccicllo. G .. Lerone. M .. Puliti. A .. Luo. Y .. Weissenbach. J" Devoto. M .. Munnieh. A" and Romeo. G. (1993) A gene for Hirschsprung disease maps to the proximal long arm of chromosome 10. Nature Genetics 4:346-350.

Makhlouf. G.M. (1985) Enteric neuropeptides: role in neuromuscular activity of the gut. Trends Phannacol. Sci. 6Z:214-218.

Martuceicllo. G .. Bicocchi. M.P .. Dodero. P .. Lerone. M .. Cirillo. M.S .. Puliti. A .. Gimelli. G .. Romeo. G" and Jasonni. V. (1992) Total colonic aganglionosis associated with interstitial deletion of the long ann of chromosome 10. Pediatr. Surg. Int. 7;308-310.

Meier-Ruge. W. (1969) New aspeets in the pathology of the hypoganglionic megacolon. Verh. Dtsch. Ges. Pathol. 53:237-239.

Meier-Ruge. W. (1971) Ueber ein Erkrankungsbild dcs Colon mit Hirschsprung-Symptomatik. Verh. Dt"ch. Ges. Pathol. 55:506-510.

Meneghelli. U. (1985) Chagas' disease: a model of denervation in the study of digestive tract motility. Braz. J. Med. BioI. Res. 18:255-269.

Michna. B" McWilliams. N.B .. Krummel. T.M .. Hartenberg. M.A .. and Salzberg. A.M. (1988) Multifoeal ganglioneuroblastoma coexistent with total colonic aganglionosis. J. Pediatr. Surg. 23:57-59.

Miyagawa-Tomita, S .. Waldo. K. Tomita. H .. and Kirby. M.L. (1991) Temporospatial study of the migration and distribution of cardiac neural crest in quail-ehick chimeras. Am. J. Anat. 192:79-88.

Mullig:m. L.M" Kwok. J.B'! .. Healey. C.S .. Elsdon. M.!" Eng. C .. Gardner. E .. Love. D.R.. Mole. S.E .. Moore. J.K .. Papi. L" Ponder. M.A .. Telenius. H .. Tunnaeliffe. A .. and Ponder. B.A.J. (1993) Genn-line mutatiOn$ of the RET proto-oncogene in mUltiple endocrine neoplasia type 2A. Nature 363;458-460.

Newgreen. D.F .• Jahnke. L Allan. U .. and Gibbins. I.L. (1980) Differentiation of sympnthetic and enteric neurons of the fowl embryo in grnfts to the chorio-allantoic membrane. Cell Tissue Res. 208:1-19.

Newgreen. D.F .. and Erickson. C.A. (1986) The migrntion of neural crest cells. Int. Rev. Cyto!. 103:89-145.

Nishijima. E .. Meijers. C .. Tibbocl. D .. Luider. T.M .. Peters-van der Snnden, MJ.H .. van der Kamp. AW.M .. nnd Molenaar. J.C. (1990) Formation and malformation of the enteric nervous system. J. Pedint. Surg. 25:627-631.

Noden. D.M. (1988) Interactions and fates of avinn craniofncinl mcsenehyme. Development 103 Supp!.:121-140.

Nolf. P. (1934) Les nerfs extrinsiques de l'intestin chez l'oiscau. Le nerf de Remak. Arch. 1m. Physio!. 39:227~256.

Okamoto. E .. and Veda. T. (1967) Embryogenesis of intramurnl ganglia of the gut and its relation to Hirsehsprung's disease. J. Ped. Surg. 2:437-443.

Pass:ll'ge. E. (1967) The genetics of Hirschsprung's disease. Evidence for heterogeneous etiology and a study of sixty-three fnmilies. New Eng!. 1 Med. 276:138~143.

Passarge. E. (1983) Hirschsprung disease and other developmental defects of the gastrointestinal tract. In: Hirschsprung disease ami other devewpmental defects of the gastrointestinal tract. A.E.H. Emery and Dol. Rimoins (cd). Churchill Livingstone: Edingburgh.

Payette. R.F .. Bennet. G.S .. and Gershon. M.D. (1984) Neurofi1nment expression in vngal neural crest-derived precursors of enteric neurons. Dcv. BioI. 105:273-287.

60

Payette. R.F.. Tennyson. V.M .. Duc Pham. T .. Mawc. C.M .. Pomeranz, RD .. Rothman. T.P .. and Gershon. M.D. (1987) Origin and morphology of nerve fibres in the aganglionic colon ofthc lethal spotted (ls!Is) mutant mouse. 1 Compo Neurol. 257:237~252.

Pomeranz. H.D .. and Gershon, M.D, (1990) Colonization of thc avian hindgut by cells derived from the sacral neural crest. Dev. BioI. 137:378-394,

Pomeranz. H.D .. Rothman. T.P .. and Gershon. M.D. (1991a) Colonization of the post-umbilical bowel by cells derived from sacral neural crest: direct tracing of cell migration using an intercalating probe and a replication~deficient retrovirus. Development 111:647-655.

Pomeranz. H.D .. Sherman. D,L .. Smo.lheiser. N.R .. Tennyson. Y.M .. and Gershon. M.D. (l991b) Expression of a neurally related laminin binding protein by neuroJ ercst~derived cells that colonize the gut: relationship to the formation of enteric ganglia. 1 Compo Neurol. 313:625~642,

Rothman. T .. and Gershon. M,D. (1982) Phenotypic expression in the developing murine enteric nervous system. 1 Neurosei. 2:381-393.

Rothm:m, T,P .. Nilaver. G., and Gershon, M.D. (1984) Colonization of the developing murine enteric nervous system and subsequent phenotypic expression by the precursors of peptidergic neurons. 1 Compo Neurol. 225:13-23,

Rothman, T.P .. Tennyson. Y.M .. :md Gershon, M.D. (1986) Coloniz:J.tion of the bowel by the precursors of enteric glia: studies of normal and congenitally aganglionic mutant mice. 1. Compo Neurol. 252:493~506.

Rothman. T.P .. Tennyson. V.M .. and Gershon, M.D. (1987) Loss of the capacity of intrinsic neural precursors to express a catechoJaminergic phenotype is eorrel:J.ted with invasion of the developing chick gut by sympathetic nerves. Neurosci. Abstr, 13:925,

Ryan. E,T" Ecker, 1.L .. Christakis. N.A .. and Folkman. J. (1992) Hirschsprung's disease: associated abnormalities and demography. 1 Pediatr. Surg, 27:76-81.

Sadaghiani. B .. and Thicb:J.ud, C.H. (1987) Neural crest development in the Xenopus laevis embryo. studied by interspecific transplantation and scanning electron microscopy, Dev. BioI. 124:91-110.

Scldenrijk, CA. van der Harten. HJ .. KWck. P .. Tibboel. 0 .. Moonn:m-Voestermans. K .. and Meijer, C.J. (1986) Zonal agangIionosis. An enzyme and immunohistochemical study of two cases. Virehows Arch. A 410:75-8\.

Scrbedzija. G,N .. Burgan. S .. Fraser. S.E .. and Bronner-Fraser. M, (1991) Vital dye labeling demonstrates a sacral neural crest contribution to the enteric nervous system of chick and mouse embryos. Development 111:857-867,

Seymour. M.K .. Wright. K.A., and Doncastcr, c.c. (1983) The action of the anterior feeding apparatus of Caenorhabdiris degan.l· (Nematoda: Rhabditid:J.). J. Zoo!. 201:527-539.

Smith, 1.. Coehard, P .. .c:.nd LeDouarin, N.M. (1977) Development of choline acetyltransferase and cholinesterase nctivitics in cnteric ganglia derived from presumptive :J.drenergic and cholinergic levels of neuml crest Cell Diff. 6:199-216.

Smith-Thomas. L.C .. Dnvis, lP" and Epstein. M,L. (1986) The gut supports ncurogenic differentiation of periocui:J.r mesenchyme, a ehondrogcnic neural crest-derived eell popui:J.tion. Dev. Bio!. 115:293~300.

Sparkes. R$ .. Sparkes. M.C .. Kalina. RE .. Pagon. RA .. Salk, DJ .. and Disteehe. C.M. (1984) Separ:l.tion of retinoblastoma and esterase D loci in a patient with sporadic retinoblastoma and del(13) (q14.1q22.3). Hum. Gcnet. 68:258-259.

Spougc, D .. and Baird, PA (1985) Hirschsprung's disease in a large birth cohort, Teratology 32:171~177.

Swenson. 0" Rheinlander, H.F .. and Diamond, 1. (1949) Hirsehsprung's disease: a new concept of the etiology. N. Engl. 1. Med. 241:551-555.

Swenson, 0 .. Shcrman, J.O .. and Fisher. 1.H, (1973) Diagnosis of congenital megacolon: an analysis of 501 patients, J. Pedi:J.tr, Surg. 8:581-593.

Tagushi, T .. Tanaka. K" Ikec1:J.. K" and Hata. A. (1983) Double zonal aganglionosis with a skipped oIigoganglionic ascending colon. Z. Kinderchir, 38:312-315.

Tanner, M.s .. Smith. B .. and Lloyd, lK. (1976) Functional intestinal obstruction due to defiency of argyrophilic neurons in the myenteric plexus. Familial syndrome presenting with short small bowel. malrotation. and pyloric hypertrophy, Arch. Dis. Childh. 51:837·841.

61

Teillet, M, (1978) Evolution of the lumbo~sacral neural crest in the avian embryo: origin and differentiation of the ganglionated nerve of Remak studied in interspecific quaiJ..chick chimerae. W. Roux's Arch. 184:251~268.

Teillet. M.A .. Cochard. P .. :md LeDouarin. N.M. (1978) Relative roles of the mesenchymal tissues and of the complex neural tube-notochord on the expression of adrenergic metabolism in neural crest cells. Zoon 6:115-122.

Tennyson, Y.M" Payctte. R.F .. Pham, T .. Mawe. G.M" Rothman, T,P .. and Gershon, M.D, (1986a) Location of the neurons projecting to the agangliomc bowcl of Is/Is micc. Neurosci. Abstr. 12:847.

Tennyson, Y.M., Payette, RF" Pharo. T .. Rothman, T.P .. and Gershon, M,D, (1986b) Extracellular matrix proteins in the colonization of the gut by precursors from the neural crest: analysis of aganglionic zones in ls!1s muunt mice, Anat. Rec, 214:1331.

Tennyson, Y,M .. Pham, T., Rothman, T.P .. and Gershon, M,D. (l986c) Abnormalities of smooth muscle, basal laminae, and nerves in the aganglionic segments of the bowel of lethal spotted mice. Anat. Rec, 215:267-281.

Thicry, ),P., Duband. J.L .. and Tucker, G,C, (1985) Cell migration in the vertebrate embryo: role of cdl adhesion and tissue environment in pattern formation, Ann, Rcv. Cell BioI. 1:91·113.

Tiffin. M.E .. Chandler. L,R .. and Faber, H.K. (1940) Localized absence of the ganglion cells of the myenteric plexus in congenital megacolon. Am.), Dis. Childh. 59:1071-1082,

Tosney. K.W. (1978) The carly migration of neural crest cells in the trunk region of the avian embryo: An c1ectron microscopic study. Dev. BioI. 62:317~333,

Tucker. G.C" Ciment, G" and Thiery, J.P. (1986) Pathways of avian neural crest cell migration in the developing gut, Dev. BioI. 116:439·450.

Walker, A.W., Kempson. R.L" and Tembcrg, J,L, (1966) Aganglionosis of the small intestine. Surgery 60:449-457,

Wcbster. W.S. (1973) Embryogenesis of the enteric ganglia in normal mice and in mice that develop congenital aganglionic megacolon. ), Embryo!. Exp. Morphol. 30:573-585.

Weyns, A.A.L.M. (1988) The enteric nervous system in the ruminant stomach of the sheep (Ovis aries). Thesis, Erasmus University Rotterdam.

Yntema. C.L" and Hammond, W.S. '(1953) Experiments on the sacral parasympathetic nerves and ganglia of the chick embryo, AnaL Rec, 115:382,

Yntema, C,L .. and Hammond, W.S. (1954) The origin of intrinsic ganglia of trunk viscera from vagal neural crest in the chick embryo, J. Compo Neurol. 101:515·541.

Yntema, C.L .. and Hammond. W.S. (1955) Experiments on the origin and development of the sacral autonomic nerves in the chick embryo. J, Exp, Zoo1. 129:375-414.

Yunis, E" Sieber, W.K., and Akers, D.R. (1983) Docs zonal aganglionosis really exist? Report of a rare variety of Hirschsprung's disease and review of the literature. Pcdiatr. Pathol. 1:33-49.

62

Chapter 3

The experimental work

3.1. Introduction to the experimental work

In this thesis, we studied the development of the enteric nervous system and the specific role

of the posterior rhombencephalic neural crest in this process. We used the avian embryo as

a model system, because it is easily amenable to experimental manipulation and we found in

our species comparison (chapters 1 and 2) that the basic principles of neural crest cell

migration and ENS development are very similar in birds and mammals. The experimental

work described in this thesis is related to three specific questions. 1) Is the neural crest

regionally specified with regard to ENS formation? 2) Which cells or tissues provide homing

and/or differentiation signals for neural crest cells in the gut? 3) What do experiments of

nature teach us about ENS development?

It is generally agreed upon that the ENS is derived from the vagal neural crest

adjacent to the first seven somites (reviewed in LeDouarin, 1982). For the most part. this

vagal neural crest lies at the level of the posterior rhombencephalon. from the level of the otic

vesicle down to the caudal boundary of somite 5. We studied. if there is a specific axial

segment within the posterior rhombencephalic neural crest which is primarily responsible for

ENS formation. The ablation experiments. described in chapter 3.2 .• showed that the neural

crest adjacent to somites 3-5 is essential for ENS formation in the colon in vivo. We found,

however. that vagal neural crest segments anterior to somite 3 were also capable of ENS

formation in the hindgut. Trunk neural crest celis, although able to colonize the gut, did not

give rise to enteric neurons or ganglia and instead differentiated into melanocytes (chapter 3.2.

and 3.5.). We also found that vagal neural crest cells which had been in prolonged contact

with the neural tube had an increased capability to form enteric ganglia, suggesting that

precursors for enteric neurons either emigrate later than hitherto assumed or require close

contact with the neural tube for a prolonged time period in order to be able to differentiate

into enteric neurons. Furthermore. we found that vagal neural crest cells, cultured for one day

in vitro were still capable of forming enteric ganglia, whereas after four days of culture they

had lost this capacity and differentiated into melanocytes, suggesting that vagal neural crest

loose their specific characteristics upon culture and start behaving similar to trunk neural crest

cells.

It has been established that the ectomesenchymal derivatives of the posterior

rhombencephalic crest are sensitive to retinoic acid (RA) both in humans (Lammer et al ..

1985) and in a number of animal species (Shenefelt. 1972; Fantel et aI .• 1977; Kamm. 1982;

Webster et aI .• 1986). but there have been no reports on an adverse effect of RA on ENS

development. Both in the posterior rhombencephalic crest and in enteric ganglia, however. RA

receptors and binding proteins have been found. Therefore, we studied the effect of RA

administration in vivo at various developmental stages, in a dose known to affect the

ectomesenchymal derivatives. but we found no disturbances in ENS formation. These results

65

give further evidence for the existence of distinct subpopulations of neural crest cells for the

various derivatives of the posterior rhombencephalic neural 'crest (chapter 3.3.).

Secondly, we studied the role of the enteric microenvironment in neural crest cell

colonization. We found that neural crest cells which had already colonized the gut were

attracted by aneural gut, whereas neural gut was not capable of attracting neural crest cells

(chapter 3.4.). We then studied the microenvironment of aneural gut in order to identify cell

types which could be involved in neural crest cell homing andlor differentiation. We found

a layer of mesenchymal cells within the submucosa of aneural gut which reacted with the

monoclonal antibody HNK-l. a pattern of immunoreactivity which we called HNK-l mode

1. We studied the expression of the HNK-1 epitope, because it was found to be present on

migrating neural crest cells (Vincent et aI., 1983) and thought to be involved in cell adhesion

(Kruse et al., 1984), and because it was also found to be expressed by other tissues at the

time they are developmentally active (Stem and Canning, 1990). Study of the origin of these

HNK-1 immunoreactive mesenchymal cells showed that these cells arise through de­

epithelialization from the splanchnic epithelium, both in the presence and absence of any form

of extrinsic innervation (Souren et al., unpublished).

Biochemical characterization of the HNK-1 immunoreactive mesenchymal cells, which

constitute approximately 10% of the total amount of celis, identified two HNK-l carrying cell

membrane glycoproteins of 42 and 44 kD (chapter 3.6.). The HNK-l mode 1

immunoreactivity in aneural gut disappeared when vagal neural crest cells colonized the gut

and formed enteric ganglia (chapter 3.5.). When trunk neural crest cells colonized the hindgut,

they differentiated into melanocytes and HNK-l mode 1 persisted.

Most of the information on ENS development, described in this thesis so far, has been

obtained through experimental embryology mainly using avian or mouse embryos. Study of

human embryology, especially those cases in which something went wrong (experiments of

nature), could provide valuable additional information on ENS development. We performed

a retrospective clinical study of patients with ENS malformations and investigated the

occurrence of associated anomalies. The results of this study, in which we also looked for

minor abnormalities and dysmorphic signs, are described in chapter 3.7. They show that the

percentage of associated anomalies occurring with the various ENS malformations (classified

according to the lengths of the aganglionic segment). are higher than reported thus far.

Furthermore, we found that with an increasing lengths of the aganglionic segment, the

percentage of associated anomalies increased. In the majority of SS-HSCR and LS-HSCR

cases in humans, aganglionosis seems to be an isolated defect (77.1 % and 54.5%

respectively). We classified the remaining SS- and LS-HSCR patients into four groups

depending on the character of their associated anomalies. In the first group. SS-HSCR was

associated with Down syndrome (7.6%). These patients were predominantly males. The

second group entailed' syndromic' cases of HSCR the incidence of which was highest in LS-

66

HSCR (20.5% versus 5.9% in SS-HSCR). The sex ratio in 'syndrontic' cases was 1 : 1 in

both SS- and LS-HSCR. In the third group, HSCR was associated with craniofacial

dysmorphisms, whereas the fourth group entailed HSCR patients with one or more anatomic abnormalities. Such a subdivision of patient groups could help considerably in the search for

the underlying, possibly genetic, defect in Hirschsprung disease.

67

DEVELOPMENTAL DYNN>UCS 196:183-194 (1993)

Ablation of Various Regions Within the Avian Vagal Neural Crest Has Differential Effects on Ganglion Formation in the Fore-, Mid- and Hindgut MAR.JO .J.H. PETERS·VA..'" DER SANDEN. MARGARET L KIRBY. ADRIANA GITTEi\"BERGER·DE GROOT. DICK TIBBOEL. MAARTEN P. MULDER. A..'W CAREL MEIJERS Department of Anatomy, J£edical College of Georgia, Augusta. Georgia rMLK.); ,'4edical Genetic Centre Rotterdam·Leiden, Departments of Pediatric Surgery (M.JHP.IJ.d.$ .. D.T .. C.M.), and Cell Biology and Genetics (MP.M.), Erasmus Uniuersity, Rotterdam, Department of Anatomy and Embryology. State UnilJersity of L,dden. Leiden (A.G.·d.G). the l'·';etherlands

ABSTRACT The vagal neural crest adjacent to the first seven somites gives rise to both gangli· onic and ectomesenchymal derivatives. Gangli· onic derivatives are the neurons and supportive cells of the enteric nervous system (ENS). cardiac. and dorsal root ganglia. Ectomesenchymal deriv· atives are cells in the cardiac outflow tract and the mesenchymal components of thymus and para· thyroids. Ectomesench)'mal derivatives are formed by a segment of the vagal neural crest, from the level of the otic vesicle dovm to the cau' dal boundary of the third somite, called the car· diac neural crest. We performed neural crest ab­lations to study regional differences within the avian vagal neural crest with regard to the forma· tion of the ENS. Ablation of the entire vagal neu­ral crest from the mid'otic vesicle dovm to the sev· enth somite plus the nodose placode resulted in the absence of ganglia in the midgut (jejunum and ileum) and hindgut (colon). The foregut (esopha' gus. proventriculus. glz7.ard, and duodenum) was normally innervated. After ablation of the vagal neural crest adjacent to somites 3-5. ganglia were absent in the hindgut. Ablations of vagal neural crest not including this segment had no effect on the formation of the ENS. We surmise that the in· nervation of the hindgut in vivo depends specifi· cally on the neural crest adjacent to somites 3-5. whereas innervation of the midgut can be accom­plished by all segments within' the vagal neural crest. The foregut can also be innervated by a source outside the vagal neural crest.

To study intrinsic differences between various vagal neural crest segments regarding ENS for· mation. we performed chorioallantoic membrane cocultures of segments of quail vagal neural an­lage and E4 chicken hindgut. We found that all vagal neural crest segments were able to give rise to enteric ganglia in the hindgut. When the neural crest of somites 6 and 7 was included in the seg· ment, we also found melanocytes in the hindgut. suggesting that this segment is more related to trunk neural crest. Furthermore. we found that the vagal neural anlage from older embryos (>18

,~ t993 WILEY·LISS, INC.

somites) showed an increased potential to form enteric ganglia. This suggests that vagal neural crest cells that have been in prolonged. contact with the neural tube in vivo. because of either late emigration or delayed migration. have an in· creased probability to form enteric ganglia. ,0 1993 WiJey·Liss.. Ine.

Key words: Neural Crest, Ablation. HNK·l, En­teric nervous system. Melanocytes. Rhombomeres. Segmentation

INTRODUCTION

Segmentation is a widely employed strategy in de· velopment. In the vertebrate head. the most prominent manifestation of segmentation is found in the hind­brain. where the cranial neural crest is associated with segmental units in the central nervous system called rhombomeres (Lumsden and Ke.l'nes. 1989; Keynes et a1.. 1990; Guthrie and Lumsden. 1991; Lumsden et a1.. 1991). The migration pathways and developmental fate of neural crest cells in the hindbrain have been studied in isotopic quail·chick chimeras (Le LieVTe and Le Douarin, 1975). by grafting of cells labelled with triti­ated thymidine (Noden, 1975). and more recently by mieroinjection of the fluorescent dye DiI (Lumsden et a1.. 1991), In this way it has been established that the neural crest of a certain rhombomere migrates to a particular pharyngeal arch to form its specific deriva· tives. Noden (1983) showed that the anterior rhomben­cephalic neural crest is already committed to a certain phenotype before the onset of migration. He trans· planted neural crest associated with rhombomeres 1 and 2, which will normally populate the first pharyn. geal arch. to the second pharyngeal arch area associ· ated with rhombomeres 3 and 4, and found that this led

Recelved February 16, 1993. Addres:; reprmt requests to Curel Meijers. M.D .. Ph.D., Department

of Cell Biology und Geneties. ErnsmllS University. P.O. Box t738, 3000 DR Rott<c>rdrun, The Nethcrlnnds.

69

PETERS·VAN DER SANDE~ ET AL.

to the formation of first arch structures within the sec­ond arch. In addition. he showed that this cephalic neu­ral crest has patterning activity. because the ectopic mandibles had a set of muscles attached to them that were derived from the second arch but resembled first arch muscles (Noden. 1988).

Recently. it was found that the cranial neural crest is not only regionally. but also temporally specified. Cra­nial neural crest cells emerging at later times have an increased probability of assuming a ganglionic rather than an ectomesenchyrnal fate (Lumsden et a1.. 1991). Temporal specification was also described for trunk neural crest (Artinger and Bronner-Fraser. 1992). It was shown that late-emigrating trunk neural crest cells are partially restricted in their developmental po­tential and mainly differentiate into melanocytes while no longer capable of giving rise to adrenergic neurons. There is, however, no clear evidence for re­gional specification of the trunk neural crest. Here mi­gration of the neural crest is largely determined by the paraxial mesoderm. Neural crest cells emerge from the neural tube in an unsegmented way and are subse­quently restricted to the anterior part of the somite (Rickmann et al.. 1985; Teillet et a1.. 1987). This seg­mented migration correlates with intrinsic differences between the anterior and posterior parts of the somites (Keynes and Stern, 1984; Stern and Keynes, 1987; Kal­cheim and Teillet. 1989).

The vagal neural crest adjacent to the first seven somites. forms a transitional zone between cranial and trunk neural crest. The vagal neural crest is generally considered to be the source for the neurons and sup­portive cells of the ENS along the entire digestive tract (Yntema and Hammond. 1954; Le Douarin and Teillet. 1973; Allan and Newgreen. 1980). Apart from this con­tribution to the ENS. the vagal neural crest also gives rise to cardiac and dorsal root ganglia and ectomesen­chymal derivatives. Dorsal root ganglia are fonned by the neural crest caudal to somite 5. Ectomesenchymal derivatives. such as cells in the cardiac outflow tract (Kirby et a1., 1983), thymic stromal cells (Bockman and Kirby, 1984). and the mesenchymal component of the parathyroids (Le Lievre and Le Douarin. 1975). and the cardiac ganglia (Kirby and Stewart. 1983) are formed by the cardiac crest from the level of the otic vesicle down to the caudal boundary of the third somite. It has been shown that there is regional specification within this cardiac neural crest (Kirby et a1.. 1985: Besson et al.. 19861. Using ablation experiments. Besson et a1. showed that the size and the location of the lesions influenced both the incidence and the type of cardiac defects. Formation of the cardiac ganglia after ablation of the cardiac neural crest could be partially rescued by the nodose placode, which proved to be capable of giv­ing rise to the neuronal derivatives of the cardiac crest (Kirby. 1988).

We studied regional differences within the vagal neural crest with regard to the formation of the ENS using two experimental systems. First. we performed

70

TABLE 1. Formation of Enteric Ganglia After Neural Crest Ablation~

Ablation n Foregut Midgut Hindgut MO-S? + placode 6 + MO-S3 = placode 5 + SI 3 + SI-2 3 + + S3-5 8 + S3-7 3 + + ... (1)

- (2)

S6-7 6 +

"The different parts of the gut were analyzed for the presence of enteric ganglia using the HNK-l antibody. + indicates normal innervation; - indicates absence of enteric ganglia. The number between parentheses indicates th(! number of embryos with or without enteric ganglia. Absence of such a number indicates that all embryos within a group gave iden­tical results.

neural crest ablation experiments to study regional dif­ferences in vivo. In a second set of experiments, we cocultured different segments of quail vagal neural crest and aneural chicken hindgut, on the chorioallan­toic membrane to study intrinsic differences between these vagal neural crest segments. Using this coculture system. we also studied whether the vagal neural crest is temporally specified with regard to the formation of ectomesenchymal and ganglionic derivatives.

RESULTS Neural Crest Albations and the Development of the Enteric Nervous System

Normal neural crest cell colonization of the chicken gut occurs between stage 19 and 32 (E3.5·E8) (Meijers et al.. 1987). We performed different types of neural crest ablations (listed in Table 1) at stage 8. and stud­ied the presence of enteric ganglia at Ell. a stage at which ganglion formation is normally completed. Ab­lations including the entire vagal neural crest from the level of the mid-otic vesicle (at the boundary between rhombomeres 5 and 6) down to the posterior boundary of somite 7 together with the nodose placode. were ex­pected to result in aganglionosis. defined as absence of enteric ganglia. in the entire gut. The nodose placode. known to be a compensatory source for cardiac ganglia after ablation of the cardiac crest. was included in the ablation to exclude possible compensation for the en­teric ganglia. We found that the esophagus. proventric­ulus. gizzard. and duodenum contained enteric ganglia. In the duodenum. enteric ganglia were observed on each side of the circular smooth muscle layer (Fig. 1). Outside the muscle layer the prominent myenteric plexus was present. whereas the submucous plexus was less well developed. Staining with hematoxylin clearly showed the presence of neurons within the enteric gan­glia characterized as large cells with a large nucleus and a clear nucleolus (Fig. 1A.B). Both plexuses were strongly immunoreactive with the monoclonal antibod­ies HNK-l and RMO 270. HNK-l stained perikarya

VAGAL NEtiRAL CREST ABLATIO::-r

G

....• ~ .. . ~:. '\

fI1'"... .'. , ' '\

"I· .'.: .~'. , ,

'~----

Miio' .

Fig. 1. Paraf'jin soctions ot Ell chicken gut aMr ablation the vagal neural cras: Irom MO-S7 at stago 10, A-F: Duodenum .. A: Hematoxylin staining shOWing tna oit/oren! layors In the duodenum. Ep: epithelium. 8m: submucosa, 8ML: smooth muscle layer. So: serosa.? ., pancreas. >( 16. B: Dotail of A, snowing myentorlc ganglia containing neurons (ar­rows). >(63. C: Immunoperoxldaso staining With HNK-l. ShOwlr'lg tn€:' myonterlc (M) and submucous (S) ganglia. >( 16. 0: Detail ot C, shOWing both neurons (arrowhoads) and extrinSIC nerve fibros (arrows) wlthm t!1e

"-I ,

f

-. ~ ,

H , ,-

myenteric ganglia. "< 63, E: Immunoperoxldase staining with RMO 270 cloany shOwing myentenc (M) and, to a lesser extent Submucous ganglia (S). x , 6. F: Dotall 01 E, showmg extrinSIC nerve Ilbres (arrows) Within tho myontenc ganglia. x 63. G: Oesopl1agus and proventriculus: the HNK·' antloody visualizes myontenc ganglia (M) and dispersed cells In the sub­mucosa (arrows). x 16. H: Gizzard: the HNK·, an~body shows myentenc ganglia (M) and dispersed staining 01 cells and libres In t!1e muscle layor and the submucosa. x, 6,

71

PETERS-VAN DERSANDEN ET AL.

and intrinsic and extrinsic nerve fibres within the en­teric ganglia (Fig. lC,D), whereas &\10 270 mainly stained nerve fibres (Fig. lE,F). In the esophagus, proventriculus and gizzard staining with the mono~ clonal antibody HNK-1 showed prominent myenteric ganglia and dispersed immunoreactive cells within the submucosa (Fig. IG,Hl. We found that the gut distal to the duodenum was aganglionic. Outside the muscle layer plexus-like structures were present, but staining v.rith hematoxylin showed that these plexuses did not contain neurons (Fig. 2A,B), Staining with the HNK-l antibody (Fig. 2C,D) revealed a layer ofHNK-l immu­noreactive mesenchymal cells in the submucosa and staining at the site of the myenteric plexus previously seen in cultures of E4 hindgut (Luider et al.. 1992). Staining with RMO 270 (Fig. 2E,F) showed the pres­ence of extrinsic nerve fibres within the myenteric plexus, but not at the site of the submucous plexus.

After ablation of the neural crest from the level of the mid-otic vesicle down to the posterior boundary of somite 3 (MO-S3), the entire gut contained enteric gan­glia (Fig. 3A,B). Including the nodose placode into the ablation did not inf1uence this result, indicating that the nodose placode did not function as a compensatory source. Ablation of the neural crest at the level of somites 3-5 or 3-7 led to aganglionosis of the colon in 10 out of 11 embryos studied (Fig. 3C.D), even in the presence of the nodose placode. In these 10 embryos there was a sharp boundary between the ganglionic and aganglionic part of the gut situated at the level of the ceca. The ceca were normally innervated, whereas the colon was aganglionic. Ablation of the neural crest adjacent to somites 6-7 had no effect on the innerva­tion of the gut (Fig. 3E,F).

The results from these ablation experiments, sum­marized in Figure 4, show that the neural crest adja­cent to somites 3-5 is essential for the formation of enteric ganglia in the hindgut. whereas the neural crest from MO to S3 and S6 to 7 is not essential for ENS formation, Furthermore. these results show that the foregut can be innervated by a source outside the vagal neural crest.

Colonization Assay

All parts of the vagal neural crest are capable of forming enteric ganglia in the hindgut. We studied whether the results obtained in our ablation experi­ments were based on intrinsic differences between the various neural crest segments regarding their ability to innervate the hindgut. The vagal neural anlage from quail embryos having 22 to 28 somites (stages 15-16) was divided into small segments and cocultured with chicken aneural hindgut on the chorioallantoic mem­brane. In Table 2 the various vagal neural crest seg­ments used in this coculture system are listed. All the segments tested were able to give rise to a normal pat­tern of enteric ganglia in the hindgut (Fig, 5). Staining with Hoechst 33258 confirmed that the cells within the enteric ganglia were of quail origin. These results in·

72

dicate that an amount of vagal neural crest equivalent to the lengths of two somites is sufficient for coloniza­tion of the hindgut. In 2 out of the six cocultures in which the posterior part of the vagal neural crest at the level of somites 4-7 was included, melanocvtes were present in the hindgut (Fig. SD). -

We conclude that with this coculture system a re­gional specification could be demonstrated regarding the neural crest adjacent to somites 6-7. ViThile all va­gal neural crest segments were able to give rise to en­teric neurons. neural crest segments including somites 6-7 in addition gave rise to melanocytes in the hind­gut.

Vagal neural crest cells from embryos of stages 13-16 show an increased potential to form enteric ganglia. To study temporal specification within the va­gal neural crest, we studied whether vagal neural crest cells from embryos of various developmental stages were equally capable of forming enteric ganglia in the hindgut. We explanted vagal neural anlagen from em­bryos having 9 to 28 somites (stages 10-16) and tested these in our coculture system. In 9 out of 10 cocultures in which the vagal neural crest was taken from an embryo with 20 to 28 somites, a normal pattern of en­teric ganglia was observed (Fig. 6Al. Of the 10 coeul­tures with vagal neural crest from embryos with 18 or less somites, only 1 showed a normal amount of enteric ganglia. In 5 coeultures with neural crest from younger embryos. enteric ganglia were present, but these were smaller, containing fewer enteric neurons, and present in less abundance (Fig. 6B), In 4 of these cocultures no enteric ganglia were present. The results are listed in Table 3,

We conclude that. in our coculture system, vagal neural crest cells from older embryos, that have been in prolonged contact with the neural tube in vivo. have an increased potential to form enteric ganglia compared to vagal n~ural crest cells from younger embryos.

DISCUSSION Regional Differences Within the Vagal Neural Crest Regarding ENS Formation

We investigated whether there are regional differ­ences within the vagal neural crest with regard to the formation of the ENS using two microsurgical ap­proaches, We found that ablation of the entire vagal neural crest from the otic vesicle down to the seventh somite resulted in aganglionosis of the mid- and hind­gut. The foregut down to the level of the duodenum was nonnally innervated. Ablation of the neural crest from the mid-otic vesicle down to somite 3 had no effect on enteric ganglia formation in the entire gut, whereas ablation of the neural crest of somites 3-5 resulted in aganglioMsis of the hindgut. These results indicate that the dependence on specific neural crest segments differs for the various parts of the gut. Innervation of the hindgut depends on a specific segment of the vagal neural crest adjacent to somites 3-S. Innervation of the

VAGAL ;'{EURAL CREST ABLAT[ON

Fig. 2. Pnrartin sedons 01 Ell chicken colon <Irter ablation of the lIagat nourol crost at tho 101101 MO-S7::1t stago 1 O. A: homat0l<Y11n staining snowing the dlfforont layers In tho colon. Ep: epithelium, Sm: submuccsa, SML: smooth musclo layor, Se: sorosa, Mes: mesentory, R: ganglion €If Aomak. x 16. B: Dotail of A, showing the neuron-froo myenteric pioxus. Note tho prosence of ono coil W1th a small nucleus, not characteristic for neurons Within tho ptexus (arrow) x 63. C: Immunepcrexldase sc<unlng with the HNK-' antibody shewing staining at tl10 site €If the myontonc

midgut, although dependent on the presence of vagal neural crest, does not depend on a specific segment. The foregut, which is normally innervated by the vagal neural crest, can also be innervated by a source outside the vagal neural crest. This might be related to intrin· sic differences between the various vagal neural crest segments in their ability to innervate different parts of the gut. This could, however, also be caused by a dif-

ploxu$ (arrows), plus an addiliOnal band 01 HNK-t immunoreactille mos­onchymal coils In tho submucosa (arrowheads). Rcmak's gangllen (R) Is also stained. x 16. D: Dotall 01 C. shOWing a myonterlc ploxus containing oxtrlnslc norvo fibres and no neurons. x 63. E: The AMO 270 antibody shows Immunoroactllllty::lt tho sito of tho myentoroc plexus (arrows), but not In tho submucoso. Remak's gongllon (R) Is alsO stained. x 16. F: Dotail 01 E. shOWing a myentorlc ploXtis contulning oxtrlnsic nerve flbras, <63.

ference in the extend of compensatory mechanisms for the various parts of the gut.

McKee and Ferguson (1984) performed unilateral or bilateral extirpation of the mesencephalic neural crest in chicken embryos and found that the mesencephalic region was repopulated by 'new' neural crest cells mi· grating from adjacent anterior or posterior neura.."(ial levels, Outflow septation of the heart depends specifi-

73

PETERS-VAN DER SANDEN ET A1.

Fig.:> Par<ltfin soctlons of E12 colon after .1blallon of part 01 tho v(lgal neural crest 0.1 stage 10. A: Ablation at the lovol 01 tho otic veslcje down to tho :;:o.';dal boundary 01 somite 3: HNK·' immunoporoxldasc staining showing tho myenteric (M) and submucous (5) ganglia. Romak's gan· glion (R) ,s also stalnod. x 40. B: Same soctlon as In A stained with homatoxylin showing onterlc ganglia. x 40. C: Ablation at the I()V<.)I 01 somltos3-5: HNK·' immunoperoxldase staining Showing a band 01 mes·

cally on the cardiac neural crest and can not be com" pensated by more anterior or posterior crest (Kirby et a1.. 1983, 1985; Besson et at., 1986). Cardiac ganglia, however, which also derive from the cardiac crest, can be formed by cells derived from the nodose placode (Kirby, 1988), These results suggest that compensatory mechanisms might vary for the different segments of the neural crest and may also depend on the specific derivatives of each segment. We found normal inner" vation of the foregut after ablation of the entire vagal

74

onchymal coils In the submucosa (arro~) and staining at the site 01 tho myenteric plexus (arrowheads). Remak's ganglion (R) IS also stained. x 40. 0: Similar seCtion as In C: staining with homatoxylin shows the absence Of emerlc ganglia. x 40. E: Ablation at tho level 01 somltes 6-7: HNK-' ,mmunoperoxldase Staining shows tho presence 01 myl)nterlc (M) (lnd submucous (5) ganglia. x 40. F: Similar SOCtlon (1$ m E. st.:1med with hematoxylin. x 40.

neural crest. This could mean that the vagal neural crest, which gives rise to enteric ganglia along the en" tire digestive tract in quail"chick chimeras (10 Douarin and Teillet, 1973), may not be the only source for en­teric ganglia in the foregut. Yntema and Hammond (1954) described aganglionosis of the entire digestive tract after ablation of the vagal neural crest. This was only the case, however, when the ablation included the anterior rhombencephalic neural crest from the otic vesicle up to the level of the fifth cranial nerve (corre-

VAGAL NEURAL CREST ,\ELATION

"-Ote C 010

I I V

CD CD []J []J OJ OJ GJ GJ

Somites QJ QJ Semites QJ QJ III 'I III

\

~i 00

JI':' '

j I,VS

[ID [ID Qli]

'"

Om

Fig. 4. Schematic drawing 5Ummar1zlng the results 01 tho various neural crest abla~ons. The bars In tM embryo Indicate the ablated seg· ment 01 (he neural crest. The P<lrts 01 the gut coloured black Indicate the extent 01 the agangliono$I$, the hatched pans are normally inn(lrvated.

[ID [ID Qli]

lilli .. ::,I'I',··,II'I· •. ·.·,I!".'I!. 81 I ~';li"1 :1,,1' Go

c.a ", 'I, ,:I!:II!

R';:!!I:

!""

Top left: ablation 01 MO·S7: top right: ablation 01 $3-5; bottOm left: ablation MO·S3~ bottom right: abla~on 56-7. Oe: osophagus. Vs: pro­ventriculus, Go: gizzard, au: duodenum. Be bile duct. J: Jejunum. Om: umbiliCUS, I: Ileum, Ca: ceca. R: rectum. Oto: otiC veSicle.

75

PETERS-VAN DER SA::fDEN ET AL.

TABLE 2. Colonization Assay With Segments of the Vagal Neural Crest"

Segment of Crest n Enteric ~o.nj:;lia Melanocytes 81-3 5 4 0 84-5 3 2 0 84-7 2 2 1 86-8 4 4 1

~The different segments of vagal neural crest arc indicated by the numbers of the adjacent somites. n indicates the number of cocultures perfonned within each gToup. Enteric ganglia were visualized with the HNK-l antlbody. The amount of cocultures showing enteric ganglia nnd/or melanocytes arc indicated.

sponding to rhombomere 2). When the ablation in­cluded only the posterior rhombencephalic crest from the otic vesicle caudad, they too found enteric neurons in the foregut. This indicates that, in the absence of the vagal neural crest, the anterior rhombencepha1ic neu­ral crest may also be capable of giving rise to enteric neurons in the foregut. Our finding that ablation of the cardiac crest does not result in disturbed ENS forma­tion, could be due to compensation by neural crest an­terior to the otic vesicle or posterior to somite 3. An­other explanation could be that this part of the vagal neural crest does not contribute to ENS formation in vivo. We performed isotopic quail-chick chimeras and found that the neural crest of somites 3-5 gave rise to enteric ganglia along the entire gut. In chimeras of the neural crest ofM0--S2. we did not find quail cells in the gut, except for one chimera in which few quail cells were present in the fore- and midgut. In all chimeras of the neural crest of MO--S2. however, quail cells were foune in the heart (unpublished results). In a similar study, Le Douarin and Teillet (1973) constructed isoto­pic quail-chick chimeras. containing various parts of the vagal neural crest. Chimeras containing the neural crest of somites 1-6 or somites 4-9 both gave rise to an ENS consisting almost entirely of quail cells. Chimeras of the neural crest of so mites 6-13 gave rise to an ENS consisting of both chicken and quail cells. The results of these various types of chimeras, combined with the data of our chimeras and ablations, strongly suggest the importance of the neural crest of somites 4 and 5 for ENS formation.

After ablation of the neural crest of so mites 3-5 en­teric ganglia were absent in the hindgut. indicating that neural crest cells anterior and/or posterior to this segment can give rise to enteric ganglia ir. the fore- and midgut, but not in the hindgut. We always found a sharp boundary between the aganglionic and the gan­glionic part of the gut, which was situated at the level of the ceca. Such a sharp boundary makes it less likely that agangHonosis in the hindgut is caused by a mere quantitative defect, that is a shortage of enteric pre­cursors following ablation. Our results indicate that innervation of the hindgut in vivo depends specifically on the neural crest adjacent to somites 3-5, and can not be compensated by other sources.

76

Regional Differences Within the Vagal Neural Crest May Be Related to Migration Pathways

When we performed cocultures of small segments of quail vagal neural anlage and E4 chicken hindgut, we found that all different segments tested were capable of forming normal enteric ganglia. In a previous study (Peters-van der Sanden et a1.. 1993), we used the same coculture system and demonstrated an intrinsic differ­ence between vagal and trunk neural crest cells in their ability to innervate the hindgut. Although both vagal and trunk neural crest cells were able to colonize the hindgut, vagal neural crest ceUs differentiated into enteric neurons, whereas trunk neural crest cells mainly differentiated into melanocytes. Since we were not able to demonstrate intrinsic differences between various vagal neural crest segments, the special fea­tures of the neural crest at the level of somites 3-5 observed in the ablation experiments must be ascribed to an in vivo process, that does not take place in the coculture system. An important difference in the co­culture system compared to the in vivo situation. is the direct association of the neural anlage and the gut, bypassing the normal migration pathways in the em­bryo. The migration pathways of anterior rhomben­cephalic and cardiac neural crest cells have been the subject of extensive investigations (Noden. 1975, 1983; Kuratani and Kirby, 1991; Lumsden et a1.. 1991: Miyagawa-Tomita et 0.1.. 1991), but migration of the posterior vagal neural crest cells has been less well studied. ~cently. using whole-mount staining with the HNK-l antibody. it has been established that car­diac neural crest cells migrate predominantly along a dorsolateral pathway on their way to the third. fourth, and sixth pharyngeal arches (Kuratani and Kirby, 1991). These crest cells form the circumpharyngeal crest (Kuratani and Kirby, 1991: Miyagawa-Tomita et aI., 1991), a compact population of neural crest cells which is formed at stage 11 and gives rise to the pha· ryngeal ectomesenchyme. Caudad to the second somite, part of the crest cell population migrates along a ven­trolateral pathway through the rostral part of the somites (Rickmann et aI., 1985: Teillet et ai., 1987). Caudad to the third somite this pathway becomes the predominant one (Bronner·Fraser, 1986). It was found that the neural crest cells adjacent to somites 4-7 are the most anterior neural crest cells that do not popu­late the pharyngeal arches lMiyagawa-Tomita et aI., 1991). This could mean that the neural crest cells that are essential for the innervation of the hindgut and that could perhaps be responsible for the innervation of the entire gut. migrate along a pathway that differs from the one followed by anterior vagal neural crest cells.

In the present study we found that, in cocultures, the neural crest adjacent to somites 6-7, besides giving rise to enteric ganglia. also led to the formation of oc­casional melanocytes in the gut. Previous coculture ex­periments showed that trunk neural crest cultured with aneuronal hindgut gives rise to melanocytes in

VAGAL YEURAL CREST ABLATION

Fig. 5. ParaHln sectiOns 01 coclJltlJres 01 parts 01 qlJail E2 nOlJral an­lago and E4 chlckon hlndglJt stainod with tM HNK-1 ",ntlboay to show the prosonce 01 myenteric (M) "'r'ld slJbmlJcous (S) ganglia. A: Neural ar'llage

the gut (Smith et a1., 1977: Newgreen et a1., 1980; Pe­ters-van der Sanden et a1., 1993" whereas vagal neural crest cells rarely gave rise to melanocytes. Our results could indicate that the neural crest of somites 6-7 should be considered trunk neural crest. It is interest­ing to note that the caudal boundary of rhombomere 8 ofthe hindbrain is thought to lie between somites 5 and 6. An additional argument that the neural crest of somites 6 and 7 should be considered trunk neural crest comes from the observation that this is the most ante­rior level at which dorsal root ganglia, which are spe­cific trunk derivatives, are formed (Lim et a1., 1987).

Temporal Specification Within the Vagal Neural Crest With Regard to the Formation of Enteric Ganglia

We found that vagal neural anlagen taken from em­bryos having 20 or more somites, were still capable of giving rise to enteric neurons. Studies using quail­chick chimeras showed that ENS precursors leave the vagal neural anlage prior to the 13 Somite stage, al­though migration sometimes lasts until after the 16 somite stage (Le Douarin and Teillet, 1973). Our re­sults could indicate that the precursors for enteric gan­glia either emigrate later than hitherto assumed, or remain in close contact with the neural tube for a pro-

~: >-~(

.1 -;~

"" ; . ' A' •• ~1

\' • , \1.

B .'

from tho level of somltos 1-3. B: Neural anlage lrom the lovel 01 somltes 4-7. C: Neural anlago from the lovel 01 somllOS 6-7. x 25.0: DetaIl 01 FlglJre 5C showlI"Ig melarlocytes (arrows). x 63.

longed time-period. In our study, vagal neural anlagen were dissected without the use of digestive enzymes, thereby possibly including neural crest cells which had already emigrated from the neural tube, but still re­mained in close contact. It is somewhat puzzling that neural crest taken from younger embryos, still contain­ing all neural crest cells, gave rise to a less than nor­mal amount of enteric ganglia. This could be related to a phenomenon described by Kirby (1989), who found that the addition of mesencephalic neural crest at the level of the cardiac crest, interfered with the develop­ment of the endogenous cardiac neural crest. It could also be that the commitment of the neural crest cells during the prolonged contact with the neural tube, de­pends on an in vivo process that can not fully occur in a coculture system. Smith et a1. (1977) and Newgreen et a1. (1989), using the same coculture system, de­scribed normal innervation of the hindgut after cocul­ture with the vagal neural crest from stage 10 embryos. From their data, however. we could not determine whether the size and the amount of enteric ganglia they found resembled our cocultures with neural crest from younger or older embryos.

The first clues to the molecular mechanisms under­lying specification within the vagal neural crest re­garding ectomesenchymal and ganglionic derivatives,

77

PETERs.VAN DER SAJ.'ffiEN ET AL.

• .. 0

Flg. S. ParaHin sections 01 cocultures 01 quail vagal neural al1lago ,md E4 chickel1 hindgut. A: Vagall10ural crest obtlil1Od from an embryo with 28 somltes: HNK-' ImmunoporoxJdase stall1li1g shows tho presOl1ce 01 a normal amount 01 myenteric (M) al1d submucous (8) ganglia. ;.: 25. B: Vagal l1eural crest obtalnod from an embryo with 18 semites: enteriC ganglia are prosel1t. but thoy are smallor and loss abun<lant than In Figure 4A. x 25.

comes from studies with transgenic mice. A knock-out mutation of the hox-l.S gene (Chisaka and Capecchi, 1991) results in a phenotype that somewhat resembles the human DiGeorge anomaly, characterized by partial or total absence of the thymus and parathyroids often combined with cardiac outflow tract anomalies. All these defective organs receive an ectomesenchymal contribution from the vagal neural crest. A knock-out mutation of the Mx-l.6 gene (Lufkin et al., 1991; Chisaka et aI.. 1991) specifically affects the neurogenic crest of the hindbrain, whereas overexpression of the Mx-ZA gene results in ENS defects (Wolgemuth et al., 1989; Gershon and Tennyson, 1991). Therefore, within the same region of the hindbrain (rhombomeres 4-7), populations of neural crest cells (neurogenic and ecto­mesenchymal) may be differentially specified by sev­eral hox genes that belong to the same cluster, but that exhibit different spatial and temporal patterns of ex­pression.

EXPERIMENTAL PROCEDURES Animals

Fertilized chicken-Gallus gallus domestic us-and quail-Coturnix coturnix japonica-eggs were incu-

78

TABLE 3. Colonization Assay With Vagal Neural Crest From Embryos of Different Ages"

9-18 20-28

n

10 10

+ 1 9

Enteric !i;:tmglia

= 5 4 0 1

~Age of the embryos is indicated by the total number of somites (S). The amount of enteric ganglia in the cocultures is scored: .... l normal amount of ganglia as compared with in vivo. =l fewer ganglia present than in vivo. -) no enteric ganglia present. n is the total number of cocultures in each group.

bated at 38°C in a forced draught humidified incubator. Chicken embryos were staged according to the table of Hamburger and Hamilton (1951). quail embryos were staged according to their number of somites. We used quail neural primordia and chicken hindgut in our co~ culture experiments. The quail condensed heterochro~ matin was used as a marker to detect the presence of quail neural crest cells in the chicken hindgut (Le Douarin and Teillet. 1973).

Neural Crest Ablation

Embryos were incubated for approximately 30 hr un­til they reached stages 8-10. Experimental animals were stained in situ with 0.02% neutral red through a window in the shell prior to carefully tearing the vi­telline membrane in order to expose the neural folds. Portions of neural folds within the vagal neural crest region were ablated bilaterally by microcautery as has been described previously (Kirby et aI., 1983). Shams were windowed and stained, and the vitelline mem~ brane was torn, but the embryos were not altered. The windows were sealed and incubation was continued. The embryos were harvested after eleven days of incu­bation and fixed in either 4% parafonnaldehyde or 70% ethanol containing 150 mM NaCl. After fixation the embryos were dissected and the isolated gut was di­vided into proximal and distal parts.

Colonization Assay

A 1 mm piece of hindgut just distal to the cecal bulges was isolated from 4-day·old chicken embryos (stage 22(23). At this developmental stage, this part of hindgut does not contain neural crest cells, neither va­gal nor sacral (Pomeranz et a1., 1991; Luider et al., 1992). The vagal neural crest adjacent to the first seven somites was isolated from quail embryos having 9-28 somites (stages 10-16). Neural tubes containing the neural crest were dissected and freed of somites using a microscalpel. No digestive enzymes were used. The neural tubes were divided in small parts with a length equivalent to 2 or 3 somites. Each part was placed on an Immobilon P millipore filter (Schleicher and SchueH. FRG) together with a segment of hindgut. The filter was placed upside down on the chorioallantoic membrane of a 7 ~day-old chicken host embryo. Trans-

VAGAL NEURAL CREST ABLATION

plants were harvested after a 7-day coculture period and fixed in 2% paraformaldehyde in 0.1 M phosphate buffered saline (PBS) for 24 hr at room temperature.

Immunohistochemistry

Specimens were routinely prepared for paraffin em­bedding and sectioned at 5 p..m. We used the mono­clonal antibodies HNK-1 (Abo and Balch, 1981) (ATCe; TIB 200; hybridoma supernatant, undiluted), as a marker for neural crest cells and enteric ganglia, and RMO 270 (Lee et al" 1987) Ihybridoma superna­tant, diluted 1:500) as a marker for neurofilaments. These first step antibodies were incubated for one hour at room temperature. Rabbit-anti-mouse immunoglob­ulins coupled to horseradish peroxidase (diluted 1:100; Dakopatts, Denmark) and goat-anti-mouse immuno­globulins coupled to FITC (diluted 1:20; Dakopatts, Denmark) were used as second step antibodies. PBS containing 0.1% Tween 20 was used for all rinsing. The peroxidase was visualized with 0.1% 3,3' diamino­benzidine·HCI (Serva, FRG) with 0.01% H 20 Z' Endog­enous peroxidases were inhibited by a 20 min incuba­tion in methanol/hydrogen peroxide (99:1 v/v) solution. Immunoperoxidase stained sections were counter­stained with haematoxylin for 1 min. To visualize the quail condensed heterochromatin marker. sections were incubated with Hoechst 33258 (2 p..g/ml PBS) for 4 min. The sections were analyzed using a Leitz ortho­plan fluorescence microscope. Immunoperoxidase stained sections were analyzed using a Leitz orthoplan microscope and photographs were taken with a Leitz camera using Agfa ortho film (25 ISO) and a Kodak Wratten 49B fitter.

ACKNOWLEDGMENTS

We thank Dr. J.Q. Trojanowski for his kind gift of the anti-neurofiiamcnt antibodies. We thank Dr, A.W.M. van der Kamp and Dr. R.E. Poelmann for crit­ically reading the manuscript, Prof .. J.C. Molenaar for this continuous support. Mr. T. de Vries-Lentsch for photography, and Mr. M. Kuit for drawing. This study was supported by a research grant of the Sophia Foun· dation For Medical Research (grant 105l.

REFERENCES Abo. 1' .. und Bollen. C.M. 119811 A diITercoMlutlon J.nti";c·n of human

NK and K cells idt'nt'died by J. mO(lQclonal J.ntib<xiy IHt\K·lI. J. Immunol.127:1024_1029.

Allan, I.J., J.nd Now~ecn, D.F. r 1980) The orIgin and diITeren.tlation of cnt<Jric neurom; orth(' int>:~tine ot' the fowl embryo. Am. J. :"nat. 157:137-154.

Artinger. K.B .. and BronMr·Fra.. ... r. M. 119921 PJ.rtioll r('~triction in th" dev<"lopmental pot<Jntial of lat<J emi!7o.ting uvian n!)uml Cfe~t cells. Dev. BioI. 149:149_157.

~son, W.T .. Kirby. M.L .. Vun MiC'fop. M.D., and Teub,~uut, M.D. 11986) Effects of the siz!) of :C'Jion~ of th(' cardiac n.euroll crc~t o.t various embryonic ugC'~ on incidence and typi:' of cnrdiac defc"C!.::!. Circulntion 73:360-364.

Bockmnn.. D.E .. and Kirby. M.L. 11984) D<:pC'ndence of thym1.1s devel· opment on derivutive~ of the n('Ul'.ll cre~t. Sci(,nc(' 223:498_500.

Bronn<:r·Fras<:r. M. 119861 AnalYMs of the cady 81;..'1,g(·8 of trunk r'ICU·

fal crest mil'l'o.tion in avian embryos usmg the monoclonal illltibody HNK·I. Dev. BioI. 115:44_55.

Chl.'lUka. 0., und Cap~chi. M.R. 119911 Regionally re~triC"..ed devel­opmental defect:! resulting from targeted disruption of the !D.OUSC homeobox gene Hox·1.5. ;:-.J'uture 350:4;3_479.

Chkiaku, 0 .. Musci. T.8 .. and CapeeehI. M.R. (1992) DevelopmenUl.1 ddect..~ of the cur. craniul net'V~~ o.nd hindbrain resulting from tar· ge~ disruption of the mouse homcobox gen Hox-l.6. NatUl'e 355: 516_520.

Gershon. M.D .. and Tennyson. V.M. 11991) Microenvironmental fac. tor:; In the normol and abnormal development of the enterIC ncr· vous ~y~tem. In: "The MorpnogeneslS of Down Syndrome." New York: Wilcy·Liss. Inc .. pp 257-'276.

Cuthrie, S" and Lllm~d('n. A. 119911 Formation [lnd rl'!;cneration of rhombomere bo1.1ndarie~ in the developing chick hIndbrain. Devel. opment 112:221_229.

Ham.burg",r. V., o.nd HamIlton. H.L. 11951) A senes of normal stag~~ In the development of the chIck embryo, J. Morphol. 88:49-67.

Kalcheim, C .. and Teillet. M. i 19891 Coru;oquencrr.> of somite mamp­ulation on the pat~rn of dorso.l root go.nglion development. Devel­opment 106::35-93.

Keynes. RJ .• und Swrn, C.D. (984) Segmentation in the vel'tt)brate not'V01.1S system. Nature 310:786-789.

Kc'ynes. R .. Cook. G .. Do.vies,.J .. Lumsden, A" Norri~. W .. =d Swrn, C. (1990) Segmentation and the development of the vertebro.te ner­vou~ system. J. Physlo!. 84:27-32.

Kirby. M.L .. Gak T.F .. o.nd Stewart. D.E. 11983) NeUl'ul crest cdls contribute to norm.."Il aorticopulmonary septation. Science 220: 1059-106l.

Ki rby, M,L.. and SteWart, D.E. 11983) Neural cre~t OMgl.n of cardiac ganglion cells in the chick embryo: Identificatlon:.md extirpation. Dev. BioI. 97:433_443.

Kirby. M.L .. Turnage. K.L .. and HaY':!, B.M. (1985) Charact>:rizatlon of conotruncal malformatiOn:< following ablation or"cardiac" nC'uTal crest. Anat. Rec. 213:87-93.

Kirby. M.L. 119&3) Kodose placode contrtbutes autonomIC nellrOtlll to thC' heo.n in the absenc(' of cardiac neural crest. J. Neuro:;CL 8:1089-1095.

Kirby. M.L. 119s91 Pla~ticity and predetermination of me:;encephalic and trunk neuro.l crest transplan~ into the regton of the cardiuc neural crc>lt. Dev. B·LOI. 134:40'2-41'2.

Kuratani, S.C" und Kirby, M.L. 119911 Initial migrutlon nnd di~trl­bution of th" colrdiuc neural erc~t m the avian embryo: An mtl'<). duct ion to the concept of the circumpharyng"nl crc~t. Am. J. Anat. 191:215_227.

I.e Douarin, N.M" and Teillet. M. (19731 The mLgration of neural cr~st cells to the wall of th" digestivt' tract m (lV'loln embryo. J. Embryo!. Exp. Morphol. 30:31-48.

I.eC'. ::.r.-V .. Carden. M.J" Schla"pfer, W.W., and Trojanowski, J.Q. 11987i Monocion(ll o.ntLbodi<:s d.istinliUl~h :>everal differentially phosphorylutc"<i states oCthe two largest rat neuroi1lament subllnit:; INF·H and NF·}f1 o.nd demoMtrute theLr eXLlltence 'm the nonnul Mrvous system of udult rat,8. J. Neurosci. 7:3474-3488.

Le Li.:vr", C.S., and Le Douarin. N.M. 119751 Me~enchymal deriv(l. tives of the neural cr~~t: anuly~1.s of chimnerLc quuil and chick em· bryo~. J. Embryrol. Exp. Morpho!' 34:125-154.

LIm, T.M" Lunn. E.R .. Keyn"", R,J" and Stern. C.D. (1987) The dif­fering eITL"C!.::! of ~cipital and trunk ~omlt<J~ on n<:ural d~velopm"nt in th<l chick embryo. Development 100:525_533.

Lufkm. T., Dierich. A" wMeUl'. :vI" Murk. M .• and Chambon. P. (19911 Disruption of the Hox·1.6 homeobox gene results in dC'feets Ln a region COn'esponding to its r08tral domain of expressiot'i.. Cell 66:1105-1119.

Luider. T.M .. Peter:;-vun der Sanden. M.J.H .. Molenaar.J.C" Tibbod. D .• van der Kamp. A.W.M" .lad Meijeru. C. (1992) Churucterlzutioa of HNK·l antigens durin", the form(ltion of the avian entenc t'ler­"'ous ~yst"m. Development 115:561-572.

L1.1msden. A .. and Keynes. R. 11989) ~gm"nto.l pacterns of neuronal d('velcpment in the chIck hindbrnin. Namre 337:424_428.

Lumsden, A" Spraw~on. N" und Graham. A. (1991) Segmental orlg1.n and migration of neural CTl"!t cells in the hindbram region of the chick embryo. Development 113:1281-1291.

79

PETERS-VAN DERSANDEN ET AL.

McKee. G.J., and Ferguson, M.W.J. 11984) The efTeet:! of m~nceph­aiie neutal crcst cell extirpation on the development of chicken embryos. J. Anut. 139:491-512.

Meijers. J.H.C., Tibboel, Do, van der KruIlp. A.W.M., v:;m Haperen­Heut:!, C.C.M., and Molenaur.J.C. (1987) Cell division !II migratory and aggregnted Mural crest cells in the developing ):rUt: an exper. imental approach to innetvatlOn-related motility disorders of thc­J;Ut .• J. Pc-diMr. Surg. 22:243-245.

Miyagawa-Tomita, S., Waldo, K, Tomita, Ho, and Kirby. M.L. 1199lJ Temporospatinl study of thc- migration and distnbution ot" cardiac neural cr('St in quail-chick chimeraB. Am, J. Annt. 192:79-88.

Newgreen. D.Fo, Jahnke, 1., Allan, LJ .. and Gibbins. LL. 11980) Dif­ferentiation of sympathetic and enterk neurons of the fowl embryo in graft:s to ilie chorio-allantoic membrane. Cell Tissue Rca. 208:1-19.

Noden. D.M. (1975) An analysis of the migratory behavior of avian cephalic neural cr..:.st cells. Dev. BioI. 42:106_130.

Noden, D.M. (1983) The role of the neural crest in patterning of avian cmnial skeletal. connective and muscle ti~sues. Dev. BioL 96:144-16.'5.

:-roden. D.M. (1988) rnteraC~lOn~ and fat~'S of avian craniofacial mes­enchyme. Development 103 Suppl.:121-140.

Petenl-van der Sanderl. M.J.Ho, Luider. T.M., van der Kamp. A.W.Mo, Tibboel. D .. and MeIjer:>. C. (1993) Regiorud differerlceH between vanoWl axial segment:l of the avian neural creHt regarding the for­mation of enteric ganglia. DifTer~nti(ltlon . ."lccepted for publication.

80

Pomero.n;;:, H.D., Rothman. T.P .. and GeJ:'>:!hon. M.D. 119911 Coloniz."l­tlon of th~ post-umbilical bowel by celL~ derived from sacral n<,lural crest: direct tracing of cell migratl.On u:;ing an intercalating probe and a replico.tion-deficierlc retrovirus. Devc-lopment 111:647-655.

Rickmann, M .. Fo.wcHt. J" and Keynes, R.J. (1985) Th<,l migrntion of neuml crest cells and the growth of motor axons through the rostral half of the chick somIte. J. Embryol. Exp. Morph. 90:437-455.

Smith. J" Cochard, P .. arid Le Douo:rin. N.M. 11977) Developm()nt of choline o.cctyltransfermie o.nd cholincst<lrWltJ o.ctivities in enteric ganglia derived from presumptive adreMr~c and cholinergic levels ot' rleu.ral crc~t. Cell Dill 6:199-216.

Stern, C.D .. arid K"J"TIcs, R.J. (1987) lrlteractions between somite cells: the formntion and maintenancc- of s"!jment boundarl<:::\ irl the chick embryo. Development 99:261-272.

Teille~ M" Knlcheim. C" and Le Douarin, NoM. (1987) Formo.tion of the dorsal root ganglia irl the aVl.atl. embryo: Segmenml ori';'rI urld migro.tory behaVior of the neural crest progenitor cells. Dev. BloL 120:329_347.

Wolgemuth, D.J., Behrin.ger. RR, Mostoll<lr. M.P., Brinster. RL .. and Palmiter, RD. 11989) 'l'ransgenic mice overexpre~:,\lng ~"'e moUSl) homeobox-containing gene Hox-l.4 exhibit ubnormo.l gut develop­merit. Nature 337:464-467.

Yntema, C.L .. and Hammond. W.5. (1954) The originofintrim;ic gan­glia of trunk viscera from vagal neural crest in thl) chick c-mbryo. J. Compo N()<lroL 101:515-541.

CHAPTER 3.3.

Differential effect of retinoic acid on ectomesenchymal and ganglionic derivatives of the

posterior rhombencephalic neural crest in chicken embryos

Maxjo J.H. Peters-van der Sanden. Marie-Josee Vaessen, and Carel Meijers.

Abstract

Embryonic exposure to I3-cis or all-trans retinoic acid (RA) results in a spectrum of

congenital malformations in humans and in a number of animal species. These malformations

include central nervous system (eNS), craniofacial, cardiovascular. thymic and limb defects.

A common denominator for a great majority of these malformations is the rhombencephalic

neural crest. The affected outflow tract of the heart and the thymus receive a contribution

from the vagal neural crest. overlying rhombomeres 5-8. which also gives rise to the neurons

and supportive cells of the enteric nervous system (ENS). In the developing ENS. two cellular

RA binding proteins (CRABP-I and II) and a nuclear receptor for RA (RXR) are present,

which could point to a role for RA in ENS fonnation. There have been, however, no reports

neither in humans nor in animal studies of an adverse effect of RA on the development of the

ENS. We studied the effect of RA on ENS development in chicken embryos and found that

injection of IOO~ of 2.5xlO-5 M RA into the yolk-sac. at different time points in

development, did not result in ENS defects, although it did give rise to eNS and

cardiovascular malfonnations. Using a coculture system, in which vagal neural crest cells and

aneural hindgut were cultured on the chorioallantoic membrane of a seven day old host

embryo, we were able to test higher RA concentrations. which would be lethal in early

development. We found that injection of IOOfll of 5xlO'" M RA did still not result in ENS

malformations. We conclude that a concentration of RA, which has a teratogenic effect on

the ectomesenchymal component of the vagal neural crest, does not result in ENS

malfonnations. After addition of 5xlO-s M RA to in vitro cultured vagal neural crest cells,

many neural crest cells detached from the culture substrate. while others were unaffected. This

could also suggest that there are subpopulations within the vagal neural crest which differ in

their sensitivity to RA.

81

Introduction

Retinoic acid (RA) plays an important role in pattern formation in vertebrate embryos. In particular the limb and the central nervous system seem to employ RA during crucial phases

of their development. In the developing chick limb bud, endogenous RA is differentially

distributed along the anteroposterior axis (Thaller and Eichele, '87), and administration of

additional RA by implanting RA-soaked beads, induces mirror-image duplications of the digits

(Tickle et al., '82: Tickle, '91). Recent studies in chicken, mouse, Xenopus and zebrafish

embryos on the effect of RA on the CNS, showed that a relatively low dose of RA given at

mid-gastrula stages resulted in a defect in the anterior hindbrain and its associated structures

(Holder and Hill, '91; Morris-Kay et al., '91; Papalopulu et al., '91; Sundin and Eichele. '92).

A reduction of forebrain and anterior midbrain seems to be specific for Xenopus embryos

(Durston et al., '89; Ruiz i Altaba and Jessel. '91),

Teratogenicity of 13-cis retinoic acid (isotretinoin) in humans has been described by

Lammer (,85), who found that exposure to RA during the first trimester of pregnancy results

in a spectrum of congenital malformations, including cleft palate, ear defects and mandibular

underdevelopment, as well as central nervous system (CNS), heart, thymus and limb defects.

The teratogenicity of 13-cis and of its isoform all-trans retinoic acid has been confirmed in

several animal species, including mouse, rat, hamster and monkey (Shenefelt. '72; Fantel et

al., '77; Kamm, '82; Webster et al., '86).

With the exception of the CNS and limb defects, the tissues affected by RA receive

part of their mesenchymal component from the cranial neural crest (Noden, '78; Kirby et al ..

'83; Bockman and Kirby, '84), suggesting that neural crest cells could be the target involved

in RA-induced malformations. The craniofacial mesenchyme originates from the neural crest

of the mesencephalon and anterior rhombencephalon (rhombomeres 1-5) (Noden, '78). The

outflow tract of the heart and the thymus receive a contribution from the neural crest of the

posterior rhombencephalon (rhombomeres 6-8) (Kirby et al .. '83; Bockman and Kirby, '84).

The posterior rhombencephalic crest, which is also called the vagal neural crest, also gives

rise to ganglionic derivatives, such as the neurons and supportive cells of the enteric nervous

system (ENS). the parasympathetic ganglia of the heart (cardiac ganglia), and dorsal root

ganglia. The ENS constitutes the intrinsic nervous system of the digestive tract consisting of

submucous and myenteric plexuses located on either side of the smooth muscle layer (Yntema

and Hammond, '54; LeDouarin and Teillet, '73; AJlan and Newgreen, '80). In chicken

embryos, the formation of the ENS takes place between day two (E2), when neural crest cell

migration starts, and day eight (E8). when ganglion formation in the hindgut is completed

(Meijers et al., '87). Although there is evidence of a teratogenic effect of RA On the

ectomesenchymal derivatives of the vagal neural crest, there are no reports of an adverse

effect of RA on the development of the ENS. Clinicians, however. are frequently confronted

82

with the question, if exposure to RA may have been involved in ENS malformations. such

as in isolated or syndromic cases of Hirschsprung' s disease.

At the cellular level. RA enters the cell and, after reaching the nucleus, forms a

complex with specific receptors. This RA-receptor complex is subsequently able to regulate

the expression of a number of genes. The two cellular RA binding proteins (CRABP-I and

CRABP-II), present in the cytoplasm, could either function as a shuttle to transport RA to the

nucleus, or they could regulate the concentration of free cytoplasmatic RA either through

steepening of an RA gradient or through functioning as a sink protecting the nucleus from

excess RA. The latter possibility is favoured by a study in F9 teratocarcinoma cells in which

CRABP-I was shown to influence the metabolism of intracellular RA (Boylan and Gudas,

'92). Additional evidence for a protective role of CRABP-I against excess RA comes from

the fact that CRABP-I has been found to be specifically expressed in tissues that appear to

be sensitive to RA exposure during development, such as the posterior rhombencephalic crest

(Vaessen et aI., '90; Maden et aI., '91; Ruberte et aI., '91). In the gut of chicken embryos,

the presence of both an RA receptor (RXR) (Rowe et aI., '91) and CRABP-I and CRABP-II

(Maden et aI., '89; Ruberte et aI., '92) has been described. The RXR receptor and CRABP-I

have been found in enteric ganglia. CRABP-I has also been found in neural crest cells present

in the wall of the gut prior to ganglion formation (Maden et aI., '89). CRABP-II is only

expressed in the epithelium of the gut (Ruberte et aI., '92). The presence of both an RXR

receptor and CRABP in the gut suggests that RA might playa role in ENS development.

A recent study in rat embryos (Kuratani and Bockman, '92) showed that administration

of bisdiamide. an agent which gives rise to malformations similar to RA (Taleporos et al.,

'78), resulted in disturbed vagal nerve branching to the gut. Furthermore, it has been

demonstrated in hamster embryos, that almost every organ or tissue can be affected by RA,

if the embryo is treated with a specific dose and at the critical period of development

(Shenefelt, '72). Therefore, we studied whether ENS malformations could be induced by RA

in chicken embryos, using two experimental approaches. First, we performed in vivo

experiments in which we administered a teratogenic, sublethal dose of RA at different time­

points in development. Furthermore, we used a coculture system, in which we cultured the

vagal neural crest and a piece of aneuronal hindgut on the chorioallantoic membrane of a

sevcn day host embryo. With this coculture system. we tested the effect of a ten-fold higher

dose of RA. which would be lethal in vivo.

Materials and methods

Animals

Fertilized chicken -Gallus gallus domesticus- and quail -Coturnix coturnixjaponica- eggs were incubated at 38°C

in :l. forced draft humidified incubator. Chicken embryos were staged according to Hamburger and Hamilton

83

('51), qU:ll1 embryos were staged according to the number of somites.

In vivo administration of RA

A stock solution of 10-2 Mall-trans retinolc acid (Sigma. USA) was made in DMSO (dimethyl-sulfoxide: Merck,

Germany). This stock solution was diluted in phosphate buffered saline (PBS) to a working concentration of

2.5xl0-5 M. 100).l1 of this solution was injected into the yolk-sac of a chicken egg through a punctured hole in

the shell at two different time points in development: day 0 (EO or blastoderm stage) and 4 (E4) of incubation.

The hole in the shell was closed with scotch tape. Controls were injected with 100)11 of 0.25% DMSO in PBS

or with PBS alone. Stage 15 embryos were visualized through a window in the egg-shell and treated with 1.0

).lg RA in 1)112% DMSO, which was applied onto the vitelline membrane using a Hamilton syringe. After a total

of 9 or 10 days of incubation. when neural crest cell colonization is completed, embryos were harvested and

examined macroscopically. The gut was isolated and fixated in 4% paraformaldehyde in PBS, embedded in

paraffin. sectioned at 7 ).lm and prepared for immunohistochemistry.

RA treatment of cocultures of £2 quail vagal neural anlage and £4 chicken hindgut

Chorioallantoic membrane cocultures were performed as described previously (peters-van der Sanden et al., '93).

Briefly, a 1 mm segment of hindgut just distal to the caecal bulges was isolated from four-day-old chicken

embryos (E4: I-IH stage 22123). At this developmental stage. this part of the hindgut does not contain neural crest

cells (Teillet et ai.. '78: Pomeranz et al.. '91: Luider et al., '92). Vagal neural anlagen, including neural crest,

neural tube and notochord. were isolated from stage 10-12 quail embryos. The excised neural anlage was placed

on an Immobilon P filter (Schleicher and Schuell, FRG) together with the segment of hindgut. This filter was

then placed upside down on the chorioallantoic membrane of a seven-day-old chicken embryo and cultured for

seven days. 100J..lI 5x 10-4 M RA was injected into the yolk-sac of the seven-day-old host embryo immediately

at the start of the coculture or after an initial three day coculture period. when neural crest cells first start to

colonize the gut. Controls were injected with IOOJ..lI 5% DMSO.

In order to test whether the administered RA reached the site of the graft. embryonic stem cells carrying

a construct with part of the RARS promoter containing an RA response element in front of the lacZ reporter gene

(kind gift of Dr. P. van der Saag), were grafted on the chorioallantoic membrane. After 48 hours the

chorioallantoic membrane containing the grafted cells was harvested and fixated in 2% paraformaldehydelO.2%

glutaraldehyde in PBS and st:llned for B-gal::tctosidase activity (Sanes et aL. '86).

Chorioallantoic membrane cocultures were also performed with E2 chicken vagal neural primordia

isolated from embryos treated at EO with 2.5x 10-5 M RA according to the in vivo protocol (see above). After

a total coculture period of seven days, transplants were harvested and fixated in 4% paraformaldehyde in PBS.

TranspJants were embedded in paraffin. sectioned at 7 J..l111 and prepared for immunohistochemistry.

RA treatment of in vitro cultured vagal neural crest cells

Vagal neural primordia from two-day-old quail embryos were dissected and cultured as described previously

(peters-van der Sanden et al.. '93). RA was added immediately at the start of the culture in a final concentration

of 5xlO-~ or 5xlO-6 M in 0.5% DMSO. Culture medium was changed after two days of culture and a second

dose of RA was administered. Cultures were examined morphologically every day with a phase-contrast

microscope. After a four-day culture period cultures were washed, fixated in 2% paraformaldehyde in PBS for

30 min at room temperature. and stained with HNK-l (see immunohistochemistry).

84

Immunohistochemistry

We used the monoclonal antibody HNK-l. a marker for neural crest cells and enteric neurons. as primary

antibody (Abo and Balch. 'SI) (ATCC: TIB 200: hybridoma supernatant, undiluted). incubated for one hour at

room temperature. Rabbit-anti-mouse immunoglobulins coupled to horse-radish peroxidase (diluted 1:100;

Dakopatts. Denmark) were used as a second step antibody. PBS containing 0.1% Tween-20 was used for all

rinsing. Peroxidase was visualized with 0.1% 3.3·diaminobenzidine.HCI (Serva. FRG) / 0.01% H20~ in PBS.

Endogenous peroxid:lSes were inhibited by a 20 min incubation in meth::mol/hydrogen peroxide (99:1 v/v)

solution. Sections were counterstained with haematoxylin for one min. For immunoperoxidase. sections were

analyzed using a Leitz orthoplan microscope and photographs were taken using Agfa ortho film (25 ISO) and

a Kodak Wratten 49B filter.

Results

Effect of RA administration at various developmental stages in vivo

Of the 20 eggs injected with 2.5xlO-5 M RA at EO, corresponding to the blastoderm stage,

and examined at E2, 15 contained live embryos. Eight embryos did not show macroscopically

visible malformations (Fig. lA), although four showed signs of growth retardation. Of the

A B

Figure 1: Micrographs oj embryos after injection oj2.5xlO-5 M RA at EO (blastodermstage) and examined macroscopically at E2. A) Stage 11 embryos with no macroscopically visible malformations. B) Embryo of comparable age as in A; both the cranial and trunk region are smaller than in control embryos and clearly malformed. C) Embryo, whose development arrested during gastrulation and which is clearly malformed.

85

remaining seven embryos, three had arrested development during gastrulation (Fig. IB), while

four had passed through gastrulation. but showed eNS malformations in the hindbrain region

and/or disturbances in somitic segmentation in the trunk region (Fig. IG). Of the 15 eggs

injected with 0.25% DMSO at EO and examined at E2, II embryos were recovered, seven of

which were normal and four of which showed growth retardation. None of these embryos,

however. were malformed. Of the 10 eggs treated with PBS at EO and examined at E2, seven

embryos were recovered. which were of normal size and showed no malformations. Of the

20 eggs injected with RA at EO and examined at E9, only three live embryos were recovered.

The other embryos had started development. but had died prior to stage 24. The three

embryos that had survived did not show any macroscopical malformations. The gut of these

embryos showed a normal morphology with clear presence of myenteric and submucous

ganglia on either side of the circular smooth muscle layer (Fig. 2A). The ganglia were

visualized with the HNK-l antibody. a marker for early migrating neural crest cells and their

neuronal derivatives (Vincent et al .. '83; Vincent and Thiery, '84). Of the IS eggs injected

with DMSO at EO and examined at E9. seven live embryos were recovered. These looked

normal macroscopically and the gut of these embryos contained a normal ENS. The five eggs

injected with PBS and examined at E9 gave rise to One live embryo which was normal

macroscopically and had a normal ENS. Table I summarizes the results obtained at E2 and

E9.

We also examined the ENS ofE9 embryos treated with I ~g RA at E2 (stage 10-15).

the time when neural crest cells are migrating from the posterior rhombencephalic crest on

their way to their target organs. In an earlier study. these embryos were shown to have

malformed outflow tracts of the heart characterized as double outlet right ventricle (DORV)

(Broekhuizen et al .. '92). The gut of these embryos, however, showed no abnormalities and

a normal pattern of enteric ganglia was observed (Fig. 2B) (n::::3). DMSO-treated control

embryos from this study, which showed no macroscopical abnormalities in the outflow tract

of the heart, also contained a normal ENS.

Injection of 2.5xlO-5 M RA into the yolk-sac of embryos at E4, when neural crest cells

have all migrated from their site of origin and for the most part have already reached their

target organ, had no macroscopically detectable effects On survival and subsequent embryonic

development (examined at EIO). The ENS (Fig. 2C) had developed normally (n=4). These

results show that in vivo administration of RA at various time-points in development and in

a concentration that clearly affects certain aspects of embryonic development. had no effect

on ENS formation.

86

Table 1: Survival of embryos injected with 100 ~ 5xlO-' M RA prior to incubation

Treatment Examined at E2 Examined at E9

RA n=20 5 died n=20 17 died 7 malformed 3 normal 4 growth

retarded 4 normal

DMSO n=15 4 died n=15 8 died 4 growth 7 normal

retarded 7 normal

PBS n=1O 3 died n=5 4 died 7 normal 1 normal

At £2 the surviving embryos were examined macroscopically for morphological defects.

Observed malformations mainly entailed eNS defects in the hindbrain region. At £9 we

studied both morphology and the innervation of the gut.

Effect of RA on ENS formation in chorioallantoic membrane cocultures

To test higher doses of RA. we performed chorioallantoic membrane cocultures of vagal

neural anlage isolated from normal E2 quail embryos. and E4 chicken hindgut. The E7 host

embryos were injected with 100~ 5x10-l M RA or 0.5% DMSO into the yolk-sac. This high

dose of RA. which is lethal during early embryonic development, had no detectable adverse

effect on the seven-day-01d host embryo. Injection of RA was performed either at the start

of the coculture (n=3). or after an initial coculture period of 3 days (n=3). The 3 day culture

period was chosen. because at this time neural crest cells are first observed in the hindgut in

the coculture system (unpublished results). The cocultures all contained a normal amount of

enteric ganglia (Fig. 3), as did the DMSO controls. In the cocultures injected with RA after

3 days. however. the shape of the enteric ganglia differed somewhat from controls. the ganglia

being round instead of elongated.

In order to test whether the administered RA reached the graft site. we grafted

embryonic stem cells. carrying a RA response elementllacZ reporter transgene, to the

chorioallantoic membrane and cultured these for 2 days. Cells on the chorioallantoic

membrane of an egg injected with 2.Sxl0-4 M RA were intensely blue after a 5 hour staining

period. indicating lacZ gene activity. The cells were predominantly located adjacent to blood

87

Figure 2: Paraffin sections of the gut of E9 chicken embryos after treatment with 2.5xIO-5 M RA at different time-points in development, stained with the HNK-J antibody. A) Injection of RA into the yolk­sac at EO. B) Direct application of RA onto the vitellin membrane at stage 15. C) Injection of RA into the yolk-sac at E4. In all three cases, the gut contains a nonnal ENS with myenteric (M) and submucous (S) ganglia. 16x

vessels. Cells grafted to control eggs. injected with DMSO. stained lightly blue after the same

staining period. This could be caused by the endogenous RA concentration, because in vitro

experiments showed that the cells are already sensitive to 10-8 M RA, which is the

physiological RA concentration in chicken eggs. The blood vessels of control chorioallantoic

membranes not containing cells did not stain (Fig. 4).

Vagal neural primordia, isolated from the E2 chicken embryos treated with RA,

DMSO or PBS at EO according to the in vivo protocol, were used in chorioallantoic

membrane cocultures. All vagal neural primordia, including those derived from affected

embryos, were capable of giving rise to a nonnal ENS in the coculture system (Fig. 3C).

88

Figure 3: Paraffin sections of 7-day cocultures of £2 vagal neural anlage and E4 chicken hindgut, stained with the HNK-1 antibody. A) E7 host embryo injected with 2.5xIO-' M RA at the start of the coculture: the hindgut contains a normal ENS with both myenteric (M) and submucous (S) ganglia. 25x B) E7 host embryo injected with 2.5xIO-4 M RA after 3 days of coculture: myenteric (M) and submucous (S) are present, but are of somewhat abnormal shape being round instead of elongated. 25x C) Coculture with £2 vagal neural anlage isolated from embryos injected with 2.5xIO-s M RA at EO: the hindgut contains a normal amount of myenteric (M) and submucous (S) ganglia, which are of normal shape. 25x

RA affects cell-substratum adhesion of vagal neural crest cells cultured in vitro

-.

After one day of in vitro culture, RA treated quail neural crest cell cultures did not differ

from control cultures treated with DMSO. The cultures consisted of about one to two

thousand small stellate cells. which had a flattened morphology indicative of firm attachment

to the culture substratum. After two days of culture. cells became rounded and part of the

cells detached from the surface and formed small aggregates. This effect was most clearly

observed in cultures treated with 5xlO-5 M RA, but could also be observed in cultures treated with 5xl<J6 M or to a lesser extent with DMSO alone. After three days of culture even more

cells had detached, but even after five days part of the cells remained flrmly attached to the

surface. Figure 5 shows RA treated cultures fixated after 4 days and stained with the HNK-l

antibody. Both in the RA treated cultures and in the DMSO controls. the percentage ofHNK-

1 positive cells is low, which is normal for vagal neural crest cells cultured for 4 days (Peters-

89

Figure 4: Micrographs of chorioallantoic membranes of E7 chicken embryos grafted with murine ES cells carrying a RA. response elementlLacZ reporter transgene, stained for j3-galactosidase activity after 2 days of culture. A) Chorioallantoic membrane containing ES cells exposed to 2.5xlO-4 M RA: cells are mainly localized along blood vessels, where intense staining is visible (arrows). B) Chorioallantoic membrane containing ES cells not exposed to exogenous RA.: cells, again localized along blood vessels, are lightly stained (arrow). C) Chorioallantoic membrane without ES cells and without exposure to exogenous RA: there is no staining along blood vessels, but there is some aspecific staining present.

van der Sanden et al .• ·93). The cells show a variable morphology ranging from flat stellate

to bipolar. After replating in medium without RA. treated cells reattached to the surface and

looked normal, which suggests that detachment of cell was not caused by cytotoxicity of RA.

Discussion

In this study. we examined the effects of RA on ectomesenchymal and ganglionic derivatives

90

Figure 5: Micrographs of quail vagal neural crest cells cultured for 4 days in vitro tmd stained with the HNK·I tmtibody. A) Neural crest cells that remained attached to the culture substrate in the presence of2.5xIO-5 M RA. B) Neural crest cells cultured in the presence oj2.5xIO-6 M RA. C) Neural crest cells cultured in the presence of 0.5% DMSO. In all three cases the morphology of the cells is similar varying from fiat, stellate to round or bipolar. The percentage of HNK·I positive cells was very low. 16x

> .-

A -. ,

"

B

c .

'. .....

.".-

-. : "

,-J.

, . '.

': .. ,

, . ,.'

,.~"'-'.: ..... _'~ .

"

. . ,.

',' ~ . ."

\ . ',::

.'

" -. --;'-" ~ , -., ..

, , .'

. '

of the posterior rhombencephalic (vagal) neural crest. We compared the effect of RA on the

ENS, a ganglionic derivative of the vagal neural crest, with the effect on ectomesenchymal

derivatives of the vagal neural crest, particularly the cardiac outflow tract, known to be

affected by RA exposure. We found that injection of RA into the yolk-sac at the blastoderm

stage (EO) had varying effects, ranging from disturbed gastrulation, defects in the hindbrain

region, growth retardation or embryonic death. The embryos that survived until E9 were all

normal macroscopically and contained a normal ENS. This could be due to a selection process in which only those embryos that were not, or only slightly affected by the RA treatment

survived. Therefore. we performed chorioallantoic membrane cocultures of the neural anlage isolated from E2 embryos, that were clearly affected by RA treatment at EO, and E4 hindgut,

and found that vagal neural crest cells from RA affected embryos were still capable of giving

91

rise to a normal ENS. We also examined the ENS of embryos, which were treated with RA

through direct administration onto the vitelline membrane at stage 15 of development CE2).

This treatment was shown to result in cardiovascular defects in the outflow tract of the heart,

characterized as double outlet right ventricle (DORV) (Broekhuizen et al., '92), but had no

effect on ENS formation. These results show that in vivo treatment with RA in a dose which

clearly affects several aspects of embryonic development including the development of

ectomesenchymal derivatives of the posterior rhombencephalic neural crest. has no adverse

effect on ENS fonnation. There are two main differences between the two in vivo treated experimental groups.

which could have influenced the effects of RA. The first difference is the time-point at which

RA was administered. EO versus E2 (stage 15). RA given at E2 can not affect gastrulation

and CNS development which occur prior to that stage. The second difference is the way'RA

is administered to the embryo. In the embryos treated at EO, 100f.ll of 2.5xlO-' M RA

(::::2.5xlO·3 !lIDol) was injected into the yolk-sac. whereas in embryos treated at E2. 1 ~g of

RA (=3xlO" f!IIlol) was applied directly onto the vitelline membrane. This might influence

the effective RA concentration that reaches the embryo. Injection of 100f.ll 2.5xl0-s M RA

would result in an overall RA concentration of 2.5xlO-8 M. if RA is distributed evenly over

the yolk-sac (estimated egg volume: 100ml). Administration of 1 ~g of RA would result in

an overall RA concentration of 3xlO-s M. but because the RA is applied directly onto the

vitellin membrane. the effective dose reaching the embryo could be much higher. One way

to circumvent this problem is to culture early chicken embryos in vitro and adding RA to the

culture medium and thus controlling the RA concentration reaching the embryo (Sundin and

Eichele, '92).

Injection of 2.5xIO·3 J.lITIol RA into the yolk-sac of E4 embryos did not result in

malformations at E9 in our study. neither in the cardiovascular system nor in the ENS.

Cardiovascular defects (type I ventricular septum defects). however. were found after

administration of higher doses of RA to E4 embryos (Jelinek and Kistler, '81: Hart et al.,

'90), but in these studies the development of the ENS was not investigated. Administration

of this high dose of RA to E2 embrycs resulted in a percentage of embryonic death of nearly

100% (Jelinek and Kistler. '81: Hart et al., '90). In our study, we also observed a considerable

amount of embryonic death. Combining the results of embryos examined at E2 and E9

showed, however, that embryonic death did not differ significantly between the RA (55%) and

DMSO (46%) treated groups. The fact that even control injections with PBS resulted in a high

percentage of embryonic death (46%). could indicate that the injection procedure alone is

deleterious for the embryos. It has been shown that injection of a certain amount of any fluid

into the yolk-sac can be deleterious for embryonic survival (Wyatt and Howarth, '76). 100

~l appeared to be to maximum amount of fluid, that can be tolerated for embryonic survival.

In our study, PBS injection never resulted in abnormalities in surviving embryos.

92

whereas injection with either DMSO or RA resulted in growth retardation. Only RA injection,

however, resulted in visible malformations in embryos examined at E2, consisting mainly of

eNS disturbances in the hindbrain region. The fact that no malformed embryos were found

at E9 could be due to lethality of the malformations between E2 and E9. Our results indicate

that, apart from having a possible effect on embryonic survival, DMSO affects embryonic

growth. There have been some reports on an adverse effect of DMSO on embryonic

development (Wyatt and Howarth. ·76; Hart et al .• ·90). but DMSO was never found to

induce abnormalities in structures with a neural crest derivation or contribution.

In a second experimental approach, we tested the effect of RA in a coculture system.

Injection of 100~ 2.5xlO-' M RA at the start of the 7 day coculture period did not affect

ENS formation, whereas injection of the same amount of RA after an initial 3 day culture

period resulted in the presence of a normal amount of enteric ganglia, which were, however,

round instead of elongated. This might indicate that there is a difference in RA sensitivity

between the early and late phases of ENS formation. Whereas migration of vagal neural crest

cells to the gut and homing to the correct sites seem to be insensitive to RA, aggregation into

a correct pattern of enteric ganglia might be influenced by RA. This could be related to the

disturbed interaction between vagal neural crest cells and the extracellular .matrix observed

after RA exposure in vitro. The intense blue staining of the ES cells carrying an RA response

elementllacZ reporter served as proof that the administered RA reached the chorioallantoic

membrane.

RA treatment of in vitro cultured vagal neural crest cells gave similar results as RA

treatment of mesencephalic and trunk neural crest cells (Thoro good et al., '82: Smith-Thomas

et al .• ·87). These studies showed that both mesencephalic and trunk neural crest cells

responded to RA with a change in morphology and detachment from the culture substrate.

Vagal neural crest cells also responded to RA by detaching from the culture substrate and

forming small aggregates. This suggests that RA causes a decrease in cell-substrate and an

increase in cell-cell adhesion resulting in impaired cell-extracellular matrix interactions. This

effect is probably not specific for neural crest cells and occurS in all migrating cells

(Thorogood et al., '82). Although many neural crest cells detached from the surface, a

significant amount of neural crest cells remained firmly attached to the surface. This latter

population of cells could represent neural crest cells which are not sensitive to RA.

We conclude that. in chicken embryos. a teratogenic dose of2.5xI0-3 J-lIDol RA, which

induces anterior hindbrain and cardiac malformations. does not induce malformations of the

ENS at the developmental time-points tested. Furthermore, a ten-fold higher dose of RA

(2.5x 1 0-2 J-lIDol), which is lethal in vivo, does not disturb neural crest cell colonization of the

hindgut in a coculture system. The relative insensitivity to RA of the ganglionic derivatives

of the vagal neural crest could add further evidence to the hypothesis that there are distinct

subpopulations of cells within the vagal neural crest.

93

Acknowledgements

We thank Nathalie Bravenboer and Leon Verhoog for technical assistance. We thank Dr. P.

van der Saag for his kind gift of the RA responsive-reporter cell line. We thank Dr. M.P.

Mulder for critically reading the manuscript. Prof. J.e. Molenaar for his continuous support

and Mr. T. de Vries-Lentsch and Mr. M. Kuit for photography. This study was supported by

a research grant of the Sophia Foundation for Medical Research (grant#105).

References

Abo, T .. and C.M. Balch (1981) A differentiation antigen of human NK and K cells identified by a monoclono.l antibody (HNK-l). J. Immuno!.. 127:1024-1029.

Allan. J.j .. and D.F. Newgreen (1980) The origin and differentiation of enteric neurons of the intestine of the fowl embryo. Am. J. An:lt.. 157:137-154.

Bockman. D.E .. and M.L. Kirby (1984) Dependence of thymus development on derivatives of the neural crest. Science. 2-23:498-500.

Boylan. IF .. and LJ. Gudas (1992) The level of CRABP-I expression influences the amounts and types of all-trans-retinoic acid metabolites in F9 teratocarcinoma stem cells. J. BioI. Chern .• 267:21486-21491.

Broekhuizen. M.L.A .. Y.W. Wladimiroff. D. Tibboel. R.E. Poelman. AC.G. Wenink. and A.C. Gittenberger-de Groot (1992) Induction of cardiac anomalies with o.ll-trans retinoic acid in the chick embryo. Cardiol. Young. 2:311-317.

Durston. A.J.. J.P.M. Timmermans. WJ. Hage. H.FJ. Hendriks. N.J. de Vries. M. Heideveld. and P.O. Nieuwkoop (1989) Retinoic acid causes an anteroposterior transformation in the developing central nervous sytem. Nature. 340:140-144.

Fantel. AG" T.H. Shepard. L.N. Newell-Morris. and B.C. Moffett (1977) Teratogenic effects of retinoic acid in pigtail monkeys (Macaca nemestrina). Teratology. 15:65-72.

Hamburger. V .. and H.L. Hamilton (1951) A series of normal stages in the development of the chick embryo. J. Morpho!.. 88:49-67.

Hart. R.C .. P.A McCue. W.L. Ragland. K.J. Winn. and E.R. Unger (1990) Avian model for 13-eis-retinoic acid embryopathy: demonstration of neural crest related defects. Teratology. 41 :463-472.

Holder. N .. and J. Hill (1991) Retinoic acid modifies development of the midbrain-hindbrain border and affects cranio.l ganglion formation in zebrafish embryos. Development. 113: 1159-1170.

Jelinek. R .. and A. Kistler (1981) Effect of retinoic acid upon the chick embryonic morphogenetic systems. r. The embryotoxicity dose range. Teratology. 23:191-195.

Kamm. J.J. (1982) Toxicology. carcinogenicity. and teratogenicity of some orally administered retinoids. J. Am. Acad. Dermato!.. 6:652-659.

Kirby. M.L .. T.F. Gale. and D.E. Stewart (1983) Neural crest celis contributc to normal aorticopulmonary septation. Science. 220: 1 059·1 061.

Kuratani. S.C .. and D.E. Bockman (1992) Inhibition of epibranchial placode-derived ganglia in the developing rat by bisdiamine. Anat. Rec .. 233:617-624.

Lammcr. E.J .. D.T. Chen. R.M. Hoar. N.D. Agnish. P.J. Benke. J.T. Braun. C.l Cuny. P.M. Fernhoff. AW. Grix. l.T. Lott. J.M. Richard. and S.C. Sun (1985) Retinoic acid embryopathy. N. Eng!. J. Med .• 313:837-841.

Le Douarin. N.M .. and M. Teillet (1973) The migration of neural crest cells to the wall of the digestive tract in avian embryo. J. Embryo!. cxp. MorphoL. 30:31-48.

Luidcr. T.M .. MJ.H. Pcters-van der Sanden. J.C. Molenaar. D. Tibboel, A.W.M. van der Kamp. and C. Meijers (1992) Characterization of HNK-l antigens during the formation of the avian enteric nervous system. Development. 115:561-572.

94

Maden. M .. D.E. Ong. D. Summerbell. F. Chytil.:.md E.A. Hirst (1989) Cellular retinoic acid-binding protein:.md the role of retinoie ucid in the development of the chick embryo. Dev. Biol.. 135:124-132.

Muden. M .. P. Hunt. U. Eriksson. A. Kuroiwu. R. Krumluuf. :.md D. Summerbell (1991) Retinoic ueid-binding protein. rhombomeres und the neu!':)) crest. Development. 111:35-44.

Mcijers. J.H.C .. D. Tibboel. AW.M. v:.m der Kump. C.C.M. van Huperen-Heuts. und I.e. Molen:lO.I' (1987) Cell division in migrutory :.md aggregated neurul crest cclls in the developing gut: an expcrimental approach to innervution-related motility disorders of the gut. I. Pediatr. Surg .. 22:243-245.

Morris-Kay. G.M .. P. Murphy. R.E. Hill. and D.R. Davidson (1991) Effects of retinoie ucid excess on cxpression of Hox-2.9 and Krox-20 :.md on morphologicul segmentation in the hindbrain of mousc embryos. EMBO 1.. 10:2985-2995.

Noden. D.M. (1978) Interactions directing the migrution and cytodifferentiation ofavj:.m neural crest cells. In: The 8pecificity of Embryological Interactions. D. Gnrrod. cd. Ch:lpm:.m and H:lll. London. 4-49.

Pupalopulu. N .. I.D.W. Clarke. L. Bradley. D. Wilkinson. R. Krumlauf. :lnd N. Holder (1991) Retinoi:;: ueid c:luses abnormal development :lnd segmental patterning of the anterior hindbrain in xenopus embryos. Development. 113:1145-1158.

Peters-v:.m der Sanden. MJ.H .. T.M. Luider. A.W.M. van der Kamp. D. Tibboel. and e. Meijers (1993) Regionul differenceS between various axi:J...l segments of the avi:.m neural crest regarding the formation of enteric ganglia. Differentiation. 53: 17-24.

Pomeranz. H.D .. T.P. Rothm:.m. and M.D. Gershon (1991) Colonization of the post-umbilical bowel by cells derived from sucral neural crest: dircct tracing of cell migration using an interc:J...lating probc and a replication-deficient retrovirus. Development. 111 :647-655.

Rowe. A .. N.S.C. Eager. :.md P.M. Brickell (1991) A member of the RXR nuclear receptor family is expressed in neural-crest-derived cells of the developing chick peripheral nervous system. Development. 111:771-778.

Ruberte. E .. ? DolI~. P. Chnmbon. and G. Morris-Kay (1991) Rctinoic acid receptors and cellular retinoid binding proteins II. Their differential pattern of transcription during early morphogenesis in mouse embryos. Development. 111 :45-60.

Ruberte. E .. Y. Friederich, G. Morris-Kay.:.md P. Chambon (1992) Differenti:J...l distribution patterns ofCRABP I and CRAB? II transeript.~ during mouse embryogenesis. Development. 115:973-987.

Ruiz i Alt.:lb:l. A .. :.md T.M. Jessel (1991) Retinoic acid modifies the p:lttern of cell differentiation in the centr:ll nervous system of neurula stage Xenopus embryos. Development. 112:945-958.

S:.mes. J.R .. J.L.R. Rubenstein. and J.F. Nicolas (1986) Usc of a retrovirus to study post-impl:.mtation cell lineage in mouse embryos. EMBO J .. 5:3133-3142.

Shenefelt. R.E. (1972) Morphogenesis of m:J...lformations in hamsters caused by retinoic acid: Relation to dose and st.:lge at treatment. Teratology. 5:103-118.

Smith-Thomas. L .. 1. Lott. and M. Bronner-Fraser (1987) Effects of isotretinoin on the beh:lvior of neural cre!'>t cell!'> in vitro. Dev. BioI.. 123:276-281.

Sundin. O .. :.md G. Eichelc (1992) An early marker of axial pattern in the chick embryo and it.~ respeeification by retinoie acid. Development. 114:841-852.

TaIcporos. P .. M.P. Salgo. and G. O!'>ter (1978) Teratogenic action of bis(dichloroaectyl)diamine. Ter:ltology. 18:5.

Teillet. M.A .. P. Coehard. :.md N.M. Lc Douarin (1978) Relative roles of the mesenChymal tissues :.md of the complex neural tube-notochord on the expression of adrenergic metabolism in ncural crest cells. 2oon. 6: 115·122.

Th:J...lIcr. C..:.md G. Eichele (1987) Identification and spatial distribution of retinoids in the developing chick limb bud. Nature. 327:625-628.

Thorogood. P .• L. Smith. A. Nicol. R. McGinty. and D. GaITod (1982) Effects of vitamin A on the behavior of migratory neural crest cells in vitro. J. Cell Sci .. 57:331-350.

Tickle. C .. B. Alberts. L. Wolpert. and J. Lee (1982) Local application of retinoic acid to thc limb bud mimics the action of the polarizing region. Nature. 296:564-565.

Tickle. C. (1991) Retinoic acid and chick limb bud development. Development. Supp!. 1:113-121.

95

Vaessen, M., J.H.C. Meijers, D. Bootsma, and A. Geurts van Kessel (1990) The cellular retlnok-aeid-binding protein is expressed in tissues associated with retinoic-acid-induced malfonnations, Development, 110:371·378.

Vincent. M .. J. Dub:lnd. :lnd J. Thicry (1983) A cell surface determinant expressed early on migrating avian neural crest cells. Dev. Brain Res .. 9:235-238.

Vincent. M .. and J. Thiery (1984) A cell surface marker for ncurnl crest and placodal cells: further evolution in periphernl and central nervous system. Dev. BioI.. 103:468-481.

Webster. W.S .. M.e. Johnston. EJ. Lammer. and K.K. Sulik (1986) 1sotretinoin embryopathy and the cranial neurnl crest: An in vivo and in vitro study. Dev. Bioi .. 6:211-222.

Wyatt, R.D .. and B. Howarth (1976) Effect of dimethylsulfoxide on embryonic survival and subsequent chick perfonnance. Poult. Sci .. 55:579·582.

Yntema, C.L .. and W.S. Hammond (1954) The origin of intrinsic ganglia of trunk viscer:l from vagal neural crest in the chick embryo. J. Compo NeuroL 101:515·541.

96

Colonization Characteristics of Enteric Neural Crest Cells: Embryological Aspects of Hirschsprung's Disease

By J,H. Carel Meijers, Marjo Peters van der Sanden, Dick Tibboel, Arthur W.M. van der Kamp, Thee M. Luider, and Jan C. Molenaar

Rotterdam, The Netherlands

• Th,s study explores the development 01 the enteric nervous system in avian embryos. Pllrtieular emphasis was given to <::010ni:l:3tion characteristics of neural crest celts present in primitive enterie ganglia. By eoeulturing neuronal and aneuranal gut of quail and chicken embryos, we investi· gllted if and when neural crest cells in primitive enteric gangria could detach from these ganglia, migrate, and colo­ni~e adjacent chicken gut. Quail neural crest teUs were identified using the quail nucleolar mllrker .)nd the HNK·' antibody. Enteric neurons were identified using three mono­clonal antibodies directed agains.t neurofiJament proteins.. We found that neural crest cells detached from primitive ganglia in neuronal quail gut from E6 till E9, whereas neural crest cens did not leave enteric ganglia from E10 gut. These observations sho.w that there is a transient phas.e during which enteric neural crest cells can leave the gut. To deter­mine whether neural crest cells could colonize neuronal gut we cocultured neuronal gut or the neural primordium and neuronal chicken gut (E1 1). We found that quail neural crest cells do not colonize neuronal E" gut, whereas they do colonize aneuronal gut of the same age. We suggest that aneuronal gut anracts neural crest cells by diffusing factors. Copy tight 0 1992 by W.S. Sounders Company

INDEX WORDS: Neural crest; enteric nervous system; chick embryo; Hirschsprung's disease; quail_chick chimera.

THE NEURAL CREST is a transient structure in vertebrate embryogenesis that leaves its imprint

on the entire embryo. Diseases arising from the neural crest aTe particularly diverse in clinical presen­tations, involving endocrinologic, cutaneous. neurolog­ical. digestive. pulmonary. or other types of syn­dromes)'; because neural crest cells invade almost every organ of the body.~..f) Knowledge of the pro­cesses involved in neural crest cell adhesion. migra­tion, and homing within the target organ is essential for understanding the pathogenesis of congenital malformations related to the neural crest.

Enteric neurons and their suppon:ive cells together with extrinsic nerve fibres constitute the enteric nervous system (ENS). The neural crest origin of enteric neurons was demonstrated by a variety of e),:perimental approaches.'.8 "Vagal ,. neural crest cells, adjacent to somites 1 to 7. migrate Jat.erally and fill the third and fourth visceral arches. At the level of the pharynx they disperse alongside the developing bowel through the solanchnic mesodenn.9•10 Within the bowel: neural crest cells migrate to the hindgut in a craniocaudal sequence. lJ Active craniocaudal neural

Journal of Pcdll)(rIC $Ufgqry. Vol 27. No 7 {J~)vl, 1992: pp 81 1.814

crest cell migration coincides with active cell prolifer­ation.l~ The migration of neural crest cells from the vagal neural creSt to the hindgut is the longest migratory distance known. Neural crest cell coloniza­tion of the gut correlates with HNK-l-positive mesen­chymal cells in the submucosa. 13 Neural crest cell adhesion within the enteric ganglia is mediated by N·CAL\1, Ng·CAM (the neuron-glia cell adhesion molecule), and N_cadherin. 14-16

In previous experiments. we found that at an advanced stage of development aneuronal bowel still provides migratory substrates required for active neural crest cell migration. 17 In the present study we determined whether such migratory substrates persist in neuronal bowel. In order to supply distal bowel segments with neural crest cells, newly formed neural crest cell aggregates in proximal bowel segments might well take on the characteristics of the neural crest itself in the process of cell adhesion and cell detachment. In other words. a postmigratory neural crest cell might become or provide a migratory neural crest cell. To substantiate this premise, we investi­gated whether aggregated neural crest cells in enteric ganglia of quail embryos were able to colonize aneu­ronal segments of the chick gut.

MATERIALS AND METHODS

Embryos

White leghorn chicken embryos (Galll.lS gallus domestlCl.lS) and Japanese quail embryos (COllUm): cotunw: lapon/ca) were staged nccording to thc table of Hamburger and Hamliton. J¥ pJterna­lJvely. the stage of young embryos was determined b~ counting the number of paired somlles. and the stage of older embryos was determined nccording to the number of Incubational days. Eggs were incubated at 3S·C in a forced draft incubator at a relative humidity of 80'ic.

From the Department of Paedialric Swgery and the MCC Depart. meN of Cell Bic~ and Ceneru;:;, Erasmus Uruversuy, Sophi.a Children s Hospl/.QL Roaerdam., The Nerherlands.

Presented at the 38th A=l Intemal/,()Nl( Ccrtgress of the BnllSh ASSOClallOn of Paedilurie: Swgeons. Budapes~ HLlngar)". July 24·26. 1991.

Addrcss repnrtt requests 10 l.H.C. Meliers. MD. PhD, MCC Depal1rr>.ent o/Cell Bio/cg:< and Gene:ru;:;. Erasml.lS Uniw:rslty. PO Box 1;38. 3000 DR Rotterdam. The Nerherlands.

Copynght 0 /992/r;-' W.E. Sawuiers Ccmpa.n.y 0022·3468192,' 2;Oi·0005$03.oo I 0

97

Dissection of Embryonic Gut and Grafting Experiments

Segments or embryonic sut (quail Jnd,or chick) were grafted "ione or)n comblnat1on onto the chorooallantoic membr:lne of £i chick embryos aiter abrasion of the supertic1al layer. 3t th~

bifurcation of!VIo gr~at vessels. The snf\s wer~ fu;.~d in position with :l piece of cellophane (Ciingo wr!lp sterilized 1n iO'7~ ethanol and dned in Jir). The egg was s~led "1th adhesive tape and incub:lted fot 1 week.

To determine the earliest arTiv:lI of neu~al crest cells in V:IrlOUS sezments of quaIl sut. we dis~ected from E3.5 until Ell quail embryos :lnd cultured these on the chorioallantoic membr:lne. At £4 the postumbilical bowel was divided into 3. at ES into 4. at E6 and Ei we used the 5 distal segments.

To deterr01ne whether neural crest cells 10 newly formed enteric ganglia can detach and colonize neighbounng aneuronal segmcnts. we cocultured neuronal quail gut with aneuronal chick gut. For the neuronal quaIl gut we took 11 I-mm segment. distal to the origIns of the cec;a from E6 to Ell. Chicken hindgut (the region between the ceeal pnmordia and the cloaca) was dissected from embryos beiore liS colonization by neural creSt cells.. ie. at E4 and £5. 11

To determine whether neural crest cells can colOnize neuronal gut. we cocultured neuronal quaIl gut and neuronal ch,ck gut. Neuronal quail gut was also grafted in combination wllh the vagal neural primordium. Segments of the neural tube (somue 1 to 7) were diS5ccted from:12 to 21 somite embryos (E2) using tungsten needle~. All diS5eClions were penormed in Ham's FlO I1ssue culTure medium.

Tl.Ssue Fixation

For preparing cryosections. grafts were embedded in Tissue Tek II O.c. T. Compound (Miles. Nape1">'ille. IL) and snap-frozen on iiquid nitrogen cooled isopent:lne and CryOSt:ll sectioned (10 IJ.m). Cryostat sections were mounted on microscope slides coated WIth chrome alum. Before immunocytochemieal incubation. sectl0ns were fu.ed 10 acetone rOt S minutes. and then dried on air. For preparing paraffin sections grafts were fixed in 4% formaldehyde in phosphate-buffered saline (PBS). dehydrated. embedded in par­allin. and sectioned alS to i IJ.m.

Quail neural crest cells in chick bost ~ut were ,dentified usin~ the Dt"A staining according to Feul~en and ROS5enbeck.lo Quail and chick cells could be distingu1shed on light microscopy by their different nucleolar ~tn.lctures.

Immunocytochemistry

The H."JK-I 19M antibody can be used to identIfy avian neural crest ce\Is.:o The HNK-l hybridoma cell line was purchased from the Americ;an Tissue Culture Collection TIS 200.:1 HNK-j immu­noperoxidase staining was performed on both paraffin and cryo sections using undiluted supernatant.

Three mouse monoclonal antibodies. mised asainst purified human neurofilament triplet proteins but cross-reacting to chicken neurofil:lments. were used to identifv enteric neurons. The 3G6 Ig-M antibody is spccific for the 160 ill and 200 kD NF-protein.::: the 2Fl1 antibody (Sanbio. Holland) for th~ 160 and ::'00 kD.:::l and C90 for the 200 kD. NAPA-73. a neurofilament aSSQC1ated protein of 73 kD. is present in early populations of neural crest cells.:' Supernatant contaming: the EICS IgM antibody W:lS used undi­luted.

Rabbit-anti-mouse pero~dase conjugated Immunoglobulins IDako. Copenhagen. Denmark) were used in a dilution of 1 :100. In order to reduce background stamm!;. we added ::.% chick serum to the conjugate. Peroxid= was Visualized by O.l9"c 3.3' diamlnoben­zidine 4HCI (Se1">'::I. Heidelberg. Germany) and 0.0::'<;<' hydrogen

98

MEIJERS ET AL

yo:roxide All rinSlng und diluting .... ·as done in PB$ (pH i.~) wlln 0.:<;;- Tween·:?O.

RESULTS

HNK-J-Positive Cells in Embryonic Quail Gut

We determined the moment of neural crest cell arrival in various segments of cmbryonal quail gut by studying HNK-l immunoreactivity in longitudinal sections (ES to Ell). Table 1 summarizes the HNK-l immunoperoxidase findings. At £5, dispersed HNK-1-positive neural crest cells were observed be~een the undifferentiated mesenchyme of the gizzard. HNK"l immunoreactive cells were not obServed in the midgut and hindgut. Distal to the cecum HNK-l immunoreactivity was only located in Remak's gan­glion. From £5 onward Remak's ganglion remained positive. At E6. HNK-l-positive neural crest cells were located as aggregates in the differentiating muscle layers of the gizzard. HI'<1C"l-positive neural crest cells at the site of the myenteric plexus were observed down to the umbilicus. The neural crest cells were still nOt grouped in ganglionic structures. Down to the cecum few HNK-l-positive neural crest cells were seen. At E7. HNK-l immunoreactivity visualized myenteric plexuses in the periumbilical gut: however. the submucous plexus did not show HNK-l immunoreactivity. Between the umbilicus and the cecum few HNK-l-positive neural crest cells were found. but these were not grouped in clusters of neural crest cells at the sites of the enteric ganl.";lia. The distal 2 mm of hindgut did not show HN~K-l immunoreactivity. At £8. a myenteric and submucous plexus was observed in the entire bowel.

Chorioallantoic Cultures of Embryonic Quail Gut of Various Developmental Stages

To determine whether neural crest cells were present in different segments of the gut. we cultured explants of various parts of the gut and screened for the presence of HNK-l- and neurofilament-positive enteric neurons. Table 2. summarizes the results of this culture experiment. A striking observation was that gut segments that did not contain HNK-l­positive neural crest cells at the time of explantation

Tabl~,. HNK_'-Posltlv~ Cells In SUGC"S~;V" S.,gm.,nts of Embryonic Sow.,l

80"'.' Se9menl ,. " " " '" " '"

GI~zard

Ventriculus Umbilicus , c,~

Coloreclum

Ell

ENTERIC NEURAL CREST CELL MIGRATION

Tabl~ 2. HNK'1-Positjv~ Ent"rlc N~urons In Cultur.-.d S"9m~nts oj

Embryonic Ouall Gut

Bowel S,.gmenl E' E; EO " Coloreaum 212 2/' "3 '/3 Colon 2/2 'I'

c." 7/7 212 3/3 PreCe<:al 212 'I' UmbilicllS 515 'I' '" P,eumbilical 'I' 212 '"

did cOntain HNK-l-positive enteric neurons after the cu/rore period. The HNK-l immunoreactivity in the cultures was confined to the enteric ganglia and not in the serosa or the chorioallantoic membrane. This indicates that neural crest cells do not have the tendency to leave the gut in the culture system.

Neuronal Gut Innervates Aneuronal Gut in Culcure

To determine whether neural crest cells in neuro­nal gut can detach and colonize an aneuronal hindgut segment, we cocultured aneuronal chick gut (E4) with neuronal quail gut E6 to Ell on the chorioallan­toic membrane. Table 3 shows that aneuronal chick gut (E4) comained quail cells after coculture with neuronal quail gut from E6-E8 embryos (11/11). The quail cells were localized in the enteric ganglia. HNK-l immunocytochemistry on serial sections showed that the quail cells were H.N'K-1 positive. No quail neural crest cells were found in the chick gut when neuronal quail gut (EI0) was cultured together with aneuronal chick gut (1/12). The chick gut did nOt show E/CS or NF immunoreactivity.

Neural Crest Cells Do Not Colonize Neuronal Gut

To determine whether quail neural crest cells are able to colonize neuronal gut we cultured neuronal chick hindgut Ell together with neuronal quail gut (E6). Table 3 summarizes the results of both immuno~ cytochemical analysis and the staining for the quail nucleolar marker. We never observed quail neural crest cells in the region of enteric ganglia (0/10). Special attention was paid to look for quail cells at remOte sites. Quail neural crest cells were not ob-

Table 3. CQ(:llll"'e, of Ouall Ne"ral C'est c..11 Dono<$ and ChiCK

Recipient Gilt

O<:>l"IOr au~il Neu,.1 Cre., CE~ ... Cell. tr. NeU'OM' Gul ". CEl1 ''''M

E6 10'11 0110 6113

" 11111 ,0 '0

'" 1112 '0 '0 Neural pnmo,di"m E1.5 8/8 1'10 7111

Abbnwistion: NO, nOt determ'ned.

served close to the neuronal gut segment of the chicken embryo, Apparently, quail neural crest cell do not have the capacity to colonize embryonal neuronal chicken guc.

To exclude the possibility that the invasive poten­tial of migrating neural crest cells in the bowel is less extensive~ compared to original quail neural crest cells, we cocultured neuronal chick gut (Ell) with the vagal neural primordium of quail E1.5. Only 1 of 10 neuronal bowel segments contained some quail cells after coculture.ln contrast. if the quail neural primor­dium of E1.5 or neuronal quail gut (E6) was com­bined with aneuronal bowel of 11 developmental days, quail cells were found at the sites of enteric ganglia in 7 of 11 cases (we produced this age­matched control by removing the hindgut from an chicken embryo at E4 and subsequent culturing for 7 days on the chorioallantoic membrane E4 + 7CAM). HNK-1. E/CS, and NF immunocytochemistry demon­strated that the quail cells indeed were neural crest cells or enteric neurons.

DISCUSSION

The development of the ENS provides an excellent model to study cell-cell interactions between neural crest cells and their target tissues. The present data show that quail neural crest cells m primitive enteric ganglia (E6 to ES) but not in older enteric ganglia (EI0) can colonize aneuronal segments of chick gut. Furthermore, we demonstrated that quail neural crest cells do not colonize neuronal embryonic bowel (at Ell).

The observation that dusters of neural crest cells in just neuronal embryonic bowel colonize neighbouring aneuronal segmems suggeStS the existence of dif­fusing factors that regulate cell adhesion and migra­tion of emeric neural crest cells. There is no out­grov.rth of neural creSt cells when neuronal quail gut is cultured alone on the chorioallantoic membrane. When grown together with aneuronal chick gut. a number of neural crest cells leave the neuronal quail gut and colonize the aneuronal chick gut. Therefore. we surmise that the aneuronal chick bowel produces diffusing factors that trigger the release of neural creSt cells in the neuronal segments.

The existence of such diffusing factors can also be deduced from quail~chjck chimera experiments in which the quail neural anlage is inserted into a chick host. Normally quail neural creSt cells migrate accord­ing to the migratory pathways of the new environ­ment: however, the only exception is the vagal neural crest.::1. Vagal neural crest cells that are grafted to the "adrenomedullary" neural crest (adjacent to somite 18 to 24) migrate to·the developing boweL Based on

99

this observation LeOouarin et al surmised that vagal neural crest cells differ from other neural crest cells in their reaction to diffusing factors produced by the developing gut.:::: The existence of the diffusing sub~ stances in the developing gut is also supported by the studies of Rothman et al.::5 who inserted embrvonic bowel segments in a slit aside the neural tube. They found excessive prOliferation of central neurons at the site of the tube close to the bowel. It could be that in Hirschsprung's disease such diffusing factors or their neural crest cell receptors are abnormal.

These putative factor(s) have a releasing effect on neural crest cells in neuronal quail bowel of 6, 7, and S days of development, but not on neuronal quail gut at EIO. We can only speculate about the reasons why quail neural crest cells cannot leave enteric ganglia at EIO: (1) the degree of eel! adhesion, (2) the existence of basal laminae surrounding the enteric ganglia, (3) the arreSt of neural creSt cell proliferation, (4) the acquisition of a neuronal phenotype, or (5) the absence of undifferentiated cells in enteric ganglia.

MEIJERS ET AL

The observations that quail neural crest cells were present neither in the enteric ganglia nor in the surrounding regions of neuronal chick bowel, implies that quail neural crest cells do not colonize neuronal bowel. This observation points at an interesting phenomenon concerning the migration and homing of neural crest cells in the target organ. Previously, we reported that neural crest cells still can colonize aneuronaJ bowel segments even after aneuronal bowel has been cultured for 1 or 2 weeks on the chorioalJan~ toic membraneY Neural crest cell migration does not occur at the axial level after neural crest cells have used the migratory substrates or due to the disappear~ ance of cell free spaces. but in the developing bowel. the migratory substrates remain present at least for 7 to 14 days. The migratory pathways in the embryo have been described in detail as relatively cell~free spaces lined by e".-tracel1ular matrix molecules such as fibronectin, laminin, tenascin. and cytotactin. Tucker et al did nOt find such cell~free spaces in the develop~ ing chick bowel,lO

REFERENCES

L Bol;lnde RP: The neurocristop;lthies. A unifying concept of disea~ ;lrisins III neural creSt de\le!opment. Hum P:lthoI5:409-429, 1974

:.. McCredie J: Neur:tl crest defects. A neurO:lnatomic baSIS for classifiC:ltion oi multiple m:llformations rel;lted to phocomelia. J ;.,Teufol Sci 28:373-387, 1976

3. Kissel P, Andrei JM. hcquier A: The Neurocristopathies. New York, NY. M;lS$On. 1982

4. Horstadius S: The Neural Crest. in Hall BK: The Nel.lr;ll Crest. Oxford. England. Oxford "(Jni\lefmy Press. 1988, pp 96-124

5. LeDollarin :-<"M: Thc Neural Crcst. Dmbridge. EnSland. Cambridge Uni\lersity Press, 1982

6. ;.,Toden DM: Craniofnclal development: :-Jew \liews on old problems. An3t Rec 208:1-13,1984

7. Yntema CL Hammond WS: The ori~in of intrinsic gan,!:lia of trunk viscer;l from vngal neural crest in the chick embryo. J Comp Ne\.lrol 101:S15-S41. 1954

8. LeDOllarin :-:M. T eillet MA.: The migr3tion of neur31 crest cells to the w31l of the digestive tr:let in a\lian embryo. J Embryol Exp MorphoI30:31-1S, 1973

9. Ciment G. Weston JA: Enteric neurogenesis by neural crest-derived br:lnchial 3fch mesenchymal cells. Nature 305:424-427.1983

10. Tucker GC Ciment G. Thiery JP: Pathwa;-,s of avmn neur~l creSl cell migwtion in the de\leloping gut. De\l Bioi 116:439-450. 1906

11. Allan lJ, Newsreen OF: The origin :lnd ditferenti3t1on of enteric neurons in the fowl embryo. Am J Anat 157:137·154. 1980

12. Meijers JHC, Tibboel D. van dcr !<..amp A WM, et ;11: Cell di\lision in migr310ry and aggregated neural Cfe.st cells !II the developing S\.lt. An experimental approach of innerv;Hlon-reI3ted motility disorders of the gut. J Pediatr Surg 21:243-245, 1987

13. Luider TM. Peters V;ln der Sanden MJH, Molemr.3r Jc. et al: Ch:l(acteriz.:r.tion of HNK-l 3ntigens during the formation of the ~lIlan enlenc nervous SYStem. Development (in press)

14. Thiery JP. Duband JL Rutishauser U. et al: Cell adhesion molel;ules io early chicken embryogenesis. Proc Natl AC:ld 5<:i USA 79:6737-6741. 1982

100

15. Thiery J? Delou\lee A, Grumet M. et ;II: Initial appearance and regional distribution of the neuron glia cell adheslon molel;Ule in the chick embryo. J Cell Bioi 100:442-456. 1985

16. Thiery, JP. Dub3nd J-L. Delouvee A:. The role of cen adhesion in morphogenetic movements during early embryogene­Sis. in Thiery JP, Edelm;ln GM (eds): The Cell in Contact. :-Jew York.. NY. Riley, 19bi, pp 169-196

17. Meijers JHC. Tlbboel D. V;ln der Kamp AWM, et al: The influence of the slase of differenti3tion of the gUt on the migration of neural I;ells: An experimental appro;lch of Hirschspruog's di~ase. Pediatr Res 21:132-135. 1987

18. Hamburger V. HJmilton HL: A series of normal swges in the development Ot the chick embryo. J Morphol 88:49·92. 1951

19. Feulgen R. Rossenbeck H: Mikroskopisch-chemischer N;ll;h­weis einer Nucleinsaure \10m Typu:; der Thymonnudeinsaure und die darJllf"beruhende F:lfbung \Ion Zellkernen in mikroskopis.c:hen praparaten. Hoppe xyJer Z Phy:;iol Chern 135:203-252, 1924

20. T\.lcker GC. Aoyama H. Lipinski M, et al: Identical re:"lctivi­ties of monoclonal antibodies HNK-l ;lnd NCI: Conserv3tion in \lertebrates on cell deri\led from the neural primordium and on some leukocytes. Cell Diff 14:223~230, 1984

21. Abo T. Balch CM: A differentiation Jntigen·of human N"K and K cells identified b\l a monoclonal antlbod\l (HNK-1). J Immunol 127:1024-1a:!9. 19S1 .

21. LeDou3r;n NM. Teillet M~A:. Experimenwl analy:;is oi the migration and differentiation of neurobl;l$\S of the 3UlonomiC nervous system and of neuroectodermmal mesenehymal denva­tives, IJsing J biological cell marking techni'1ue. Dev Bioi 41:162-184.1974 ~. Kluck p, van MUijen GNP, 11;10 der Kamp AWM. et al:

Hirschsprung's dise:.lSe Studied with monoclonal antineurofiiament 3ntibodies on tisslJe ~ection5, Lancet 1:652-654, 1984

24. urnent G. Weston JA: Early appe:lr;lnce in neural crest and cre$t~derived cells of antigenic determin3nl present in 311l;ln neurons .. DC'I Bioi 93:355-367. 1982

25. Rothman TP. Gershon MD. Fontaine-Perus JC, et al: The effect of back transplants of the embryonic gut wall on growth of the neuT31 tul;x:. Dev Bioi 1124:331-346. 1987

Dirrer.:ntiation (1993) S3: 17-~4

Differentiation On'OI:<flY. ="wpla.l. ,nd rnfrc.,....,d.~o" Th~'"?~

C Spring.:r-V~rla~ J993

Regional differences between various axial segments of the avian neural crest regarding the formation of enteric ganglia Marjo J.H. Petcrs~v:ln der Sanden!. Theo M. Luider l , Arthur W.:'rl. van der K:lmp 2, Dick Tibboc-Il. Carel Mc-ijcrs I

l Department 01- Pcdiatnc Surgery and ~ Medical Gen.:tics Center Rotterd.1m-Leiden, Department of Cell Biology and Genetics. Erasmus University. P.O. Box 1738. NL·3000 OR Rotterdam. The Netherlands

Accepted in revised fonn Deeember~. 1992

Abstract. The vagal neural crest adjacent to somites 1-7 gives rise to the enteric ganglia along the entin~ digestive tract. It is generally assumed that formation of enteric ganglia in preumbilical gut is independent of the axial segment from which the neural crest originates. In post~ umbilical gut. however. there is evidence that the axial segment of origin of the neural crest might be relevant to neural differentiation. In this part of the gut. we pre­viously identified a subpopulation of HNK-l·immuno~ reactive cells within the enteric mesenchyme. This immu~ no reactivity disappeared upon formation of the enteric nervous svstem. We studied the interaction between var~ ious axial segments of quail neural crest and the micro­environment in aneural chicken hindgut using chorioal­lantoic membrane cocultures. We found that neural crest cells from various axial segments could migrate into the gut and home to the correct sites. However. whereas ~agal neural crest cells differentiated into enteric neu­rons. neural crest cells from truncal segments mainly differentiated into melanocytes. The HNK-l-immunore~ activity within the enteric mesenchyme only disappeared when neural crest cell colonization was followed by dif~ fcrentiation into enteric neurons and subsequent forma~ tion of enteric ganglia. To determine whether diffcrentia~ tion of neural crest ceils in chorioallantoic membrane cocultures was influenced by the prolonged presence of the neural tube and notochord. we developed a new co~ culture system. using neural crest cells cultured in vitro. We found that the differentiation of vagal and trunk neural crest cells within the enteric mesenchyme was not influenced by the prolonged presence of the neural tube and notochord after 24 h. suggesting that there are in­trinsic differences between these neural crest cell popula~ tions. Upon prolonged in vitro culturing, the properties of vagal neural crest cells changed. and concomitantly. they lost the ability to differentiate into enteric neurons and instead differentiated into melanocytes. We con~

Correspondence 10: C. Meijers

elude that the disappearance of the HNK~l-immunore· activity within the enteric mesenchyme is correlated with formation of enteric ganglia. In our experimental system. cells from the vagal neural crest are more capable of neural differentiation in the hindgut than cells from other axial levels of the neural crest.

Introduction

The neural crest in vertebrate embryos is a transient structure extending along virtually the entire neural axis. Cells from the neural crest become \videly distributed throughout the embryo and eventually localize at specif~ ic sites and differentiate into various cell types [4. 9. 16. 24. 251. The vagal neural crest adjacent to somites 1-7 gives rise to mesectodermal and ganglionic deriva~ tivcs~ Mesectodermal derivatives. such-as cells in the car~ diac outflow tract. and the mesenchymal component of the thymus and the parathyroids. derive from a specific axial segment of the vagal neuro.l crest from the level of the otic placode down to the third somite [3. 13. 14,19]. Ganglionic derivatives of the vagal neural crest :ue the enteric ganglia [2.17,38], o.nd dorsal root gan~ glia. Dorsal root ganglia are formed only by the vagal neural crest o.djacent to somites 6 and 7. and not by crest o.djacent to somites 1-5 [33].

It has b..::en established tho.t the neurons and support~ ive cells of the enteric nervous system along the entire digestive tract derive from the vagal neural crest [2. 17. 38]. Enteric ganglia in the postumbilical gut also receive a contribution from the sacral neural crest (caudal to somite 28), but its precise fate has not yet been estab~ lished [17 .. 26.27. 30]. A number of experimental studies have shown that neural crest other than vagal or sacral is able to colonize the gut [18. 23. 32. 34]. Le Douarin and TeUlet [18] made quail-chick chimeras. in which the adrenomedullary neural crest adjacent to somites 18-24

101

was grafted to the vagal region of the neuraxis. and found quail cells alom!:-the e~tire digestive tract. In the preumbilical gut. these- quail cells had differentiated into enteric neurons. Caudad to the umbilicus. however. some quail cells were found in the wall of the gut but they were exclusively pigment cells and never participat­ed in ganglion formation. In chorioallantoic membrane cocultures of ane1.lral chicken hindgut and adrenomc­dullary neural crest. enteric ganglia were observed. but the presence of mdanocytes in the chicken hindgut was also reported (18. 23. 32. 3-1.J. These melanocytes were located on either side of the smooth muscle layer. where plexus formation normally occurs. Thus. whereas after colonization of the pn::umbilical gut o.ll trunk neural crest cells differentiate into enteric neurons. colonization of the postumbilical gut results in melanocyte differenti­ation of some trunk neural crest cells.

In a previous study, we cultured ::lUeural chicken hindgut on the chorioallantoic membrane t"or 7 days and found 0. sUbpopulation of cells within the enteric mesen­chyme that stained with the monoclonal antibody H?'-rK-1 [20]. This HNK-l-immunoreactivity was localized in the submucoso.just underneath the circular smooth mus­cle layer and between the circular o.nd longitudinal smooth musclc layer. that is at the sites where ganglion formation normally occur'). This pattern of immunore­activity in aneural gut. which we called HNK-I mode I. disappeared upon formation of the enteric nervous system. In neural gut the H?'-rK-1 antibody marked the enteric ganglia, a pattern of immunoreactivity which we called H?'-rK-l mode 2. The HNK-l antibody recognizes a carbohydrate epitope present on a large number of molecules all said to be involved in cell adhesion [II. IS]. Therefore. we speculo.ted tho.t these HNK-I~immu­no reactive mesenchymal cells hOod some role in the hom­ing of neural crest cells into the hindgut and/or their differentiation into enteric neurons. In this study. we further investigo.ted this putative role.

We studied the interaction between various axial seg­ments of the quail neural crest and the microenviron­ment or' chicken aneural hindgut by performing chorioo.l­lantoic membrane coculturcs. We confirmed that neural crest cells irom quail vagal neural anlage differentiated into enteric neurons. Crest cells from trunk neural anlage migrated into the gut and homed to the sites of the myenteric o.nd the submucous plexus. but in our experi­mental system they mainly differentiated into mclano­cytes. The HNK-J immunoreactivity within the enteric mesenchyme only disappeared when crest-cell coloniza~ tion was followed by the formation of enteric go.nglia. To determine whether differentiation of neural crest cells in our cocultures was influenced by the prolonged pres~ ence of the ne\.tral tube o.nd notochord. we developed a coculture system. in which we used neural crest cells cultured in vitro and demonstrated o.n intrinsic ditTer­ence between vagal and trunk neural crest cells.

Methods

.immail'. FertIlized chickcn (Gallus gallusdoml'sru;us) and quail (Co­lurm.~ I.'orurni.~ Japon/ca) egg~ were incubated at 38" C in a rorced-

102

draught humidiiied incubator. Embryos were staged according to the table ot' Hamburger and Hamilton [IOJ, We used quail neural primordia and chicken hindgut in our coculture exp<!riments. The q\lail 1;:0ndcnScd hl;:t<:~o<;hrQmutin murkcr W:\S I.lS<!O to o<:t'Xt the pr.:,ence of cells ,dertvd from quail ncural cr<!st in the gut [:7J.

CI)/oni=alwn tlssuy. A l·mm segment of hindgut just distal to th~ caecal bul!!cs was isolated rrom ..... d<ly·old ehiek~n embrvos (E..].: HH ,tage -::'223), At this d<!vclopme·ntal st'lge. this pu;t of the hindgut docs nOt contain either vagal or 5acro.I n<!ur,tl crest cells [20.27. 34J. ;-.Jeur~d anlag<!n, including neural crest. neural tube ;lnd notochord, were isolated t'rom 2·day-old IE2) quail embryos. V<lgal neural <Inlage adjacent to the first seven somit,-'"; wa, isobted from embryos having 12-18 somites (HH sta!:,:e j 1-13). For trunk neural anlag<!, the crest adjacent to the last 6 somites was isolated I'rom embryos huving 16-3-l. somites (HH >ID.se 12-16). The anlagen were dissected and freed of somites using a microscalpd. ~o di­gestive enzymes were used. The excised n<!ural anbge W<lS pl<lccd on an [mmobilon P filter (Schleicher ;lnd SchucH. FRG) together with the segm<!nt of hindgut. This filter was then placed upside down on the ehorioaUuntoic membrane or a i-day,old chicken embryo and cultured for 7 days. to allow quail neural cr<!st cells to colonizc the chicken hindgut,

In ~itrQ ellirure of ncural cresr cells. >.Ieural :lnlagen were isolated from 2·day·old qtwil embryos as dcscrib\:d above and placcd on a glass coverslip coated with plasma libronectin (100 ~g/ml: ob­tained I'rom h\lman plasma using S<!phudcx-gdatin). The CL1lture medium consisted of at: 1 mixture of DMEM tDulbeeeo's modi­lied Eagle's medium. Flow Laboratories. UK) and FlO medium (Gibco, USA). 15% I'etul calf serum (Sanbio. NL). 3% chicken embryo extract (prepared from II·day·old embryos), glutamine (2.9:! mg/ml), penicillin (0.75 mg/ml) and Streptomycin (1.25 mg' ml). Arter 24 h the neur:11 tubes werc scraped olT using a tungsten needle. The culture mediL1m was changed twice a wcek. After a cultL1re period varying I'rom I day to 3 wet::ks, cultures were tixed .lnd prepared for immunocytochemistry.

C%m;atil)n assay II'all neural crest t:dl~' cu{rur.:d ill I:ilro. Aftcr 16-20 h of culture the neurul anbgcn wcre semp<!d off using a tL1ngsten needle. !':cL1r:ll crest cells were harvested with trypsin EDTA (0.05%,0.02% in phosphate·butT<!red saline {PBS); WIWrv).

Ten thousand neural crest cells were seeded onto an lmmobilon P Iilter. After 30 min the cells had attached to the filter and a l-mm segm<!nt of E.t. chicken hindgL1t taken just distul from the caecal bulges was placed on top of this Immobilon P Iiller. Then the filter was placed upside down on the chorioallantoic m<!mbrane 01' a 7-day-old chicken host embryo und ClJlt\lred for 7 days.

Immunohistoch.:mistry, Transplants were narvested aft<:r the cultL1re period and lixcd in 2% pamformaldehyde in PBS lor 24 h at room tcmperature. They were embedded in par..lfftn and sectloned at 7 ~m. Cell cultures were washed twice in PBS and fixe i in 2% paraformaldehyde in PBS I'or 30 min at room temperature, We L1s<:d Ihe monoclonal antibody HNK,I as primary antibody [IJ IATCC; TIB 200; hybridoma supernatant. undiluted). incubated for I h at room temper..lture, Rabbit-anti-mouse immL1noglobL1lins coupled to FITC (diluted 1:20; DakopatL'" Denmark) or rabbit­anti-mouse immunoglobulins COL1plcd to horseradish peroxidase rdiluted 1: 100; Dakopatts, Denmark) were used as second-step antibodies. PBS containing 0,1 % Tween-20 was used for:lll rinsing. P<:roxidasc was visuali7.ed with 0.1% 3.J'diaminobenzidlne HCI [Serva. FROliO.Ot % H~O: in PBS. Endogenous pero);idases were inhibited by a 20-min incubation in methanol/hydrogen peroxide (99: J vlv) solution. S<!Ccions wen: counterstained with haem:ltoxy. lin for I min, To visualize the qLU\il condensed heterochromatin n1:lrker. sections were incL1bated with Hocchst 33258 (2 )is-'ml PBS) for .. min. For immunofluorescence, sections were analyzed using a Lciu orthoplan fluores<:ence microscope. Photographs were tak­<!n with a Leitz 35-mm camero using Kodak Ektachrome film 1400 [SO). For immunopcroxidase. sections were analyzed tlSing:l Leitl

F'~.' \A. R. Paraffin ,ection.of u. 7-day coculture of E2 quail vaS:J.1 n.:ural anlage :J.nd E4 chicken hindgut. A Immunof1uon::sc.:nee staming with the H;..fK-1 antibody ,howing a myenteriC plexus. S Double-staining with Ho.:chst 33258 showin~ the preSence .of qL1uil cells wlthm this myenteric p\e:ws i<ll'ro,..Ju?ads). x 100

Orthcpbn microscope and photcgraphs were taken with a leit;:: 35·mm camera using Ag]'a Onho lilm (25 ISO) and a Kcdak Wrat­ten 498 filter.

Results

Neural crest celfs from carious axial segments coloni::e £4 chicken hindgut, hur only uagal neural crest cells form enteric ganglia

Vagal neural anlage. We performed cocultures of quai! vagal neural anlage and E4 chicken hindgut on the chor­ioallantoic membmne (n = S). We observed clusters of quail cells in the gut at the site ot' both the myenteric and the submucous plexus as shown with the qllail het­erochromatin marker visualized with Hoechst 33258 (Fig. 1). In all cocultures with vagal neural anlage. the H:JK-l antibody revealed enteric ganglia. consisting of quail cells (Fig. 2A). The enteric mesenchyme showed no H~K-l immunoreactivity.

Trunk neural anlage. Neural anlage adjacent to the lo.st 6-8 somites was isolated from quail embryos with 16-34 somites and cocultured with E4 hindgut on the chorioal­lantoic membrane. This included neural crest from dif· ferent axial segments in the trunk region. including ad­renomedullary crest. These different trunk: crest seg­ments gave identical results. After 7 days of coculture with E4 hindgut, the HNK-I antibody did not reveal enteric ganglia (n=8). Instead. we still observed HNK~l mode 1 immunoreactivity within the enteric mesen­chyme. localized at the site of the myenteric and the submucous plexus. We observed numerous HNK-l-neg­ative melanocytes at the sites of the myenteric and the submucous plexus and a few HNK-I-immunoreactive quail cells. which could have been enteric neurons (Fig. 2B. C). These HNK-l-immunoreactivc quail cells_ however. were dispersed and had not formed enteric ganglia. It was difticult to detennine the species origin of the melanocytes. because of the large number of pig­ment granu\cs present in the cclls. Clear quail origin was demonstrated in about 10% of the melanocytes.

In control cultures of aneural hindgut without neural crest cells. melanocytes were never observed (n = 20; data not shown).

These results confinn that vagal neural anlage is ca­pable of fonning enteric ganglia in the postumbilical gut. Trunk neural crest cells arc ::tble to migrate into the gut ::tnd colonize the gut at the appropriate sites. but they are not able to fonn enteric ganglia. and m::tinly differentiate into melanocytes. HNK-l mode 1 immuno­reactivity only disappeared when crest cell colonization was followed by the formation of enteric ganglia.

DifferentiatiOn within the enteric microent'ironmcnt is determined hy intrinsic properties of vagal and mink neurat crest cells and nOt influenced h.1! the prolonged presence of the neural tuhe and notochord

We studied whether the variation in differentiation ob­served between vagal and trunk neural anlage was in­tluenced by the prolonged presence of the neural tube :lnd notochord during the colonization assay. or based on an intrinsic difference between vagal and trunk neural crest cells. We developed a new coculture system. in which neural crest cells we[l~ allowed to migrate out of the neural tube in vitro. After 24 h the neural crest cells were collected, seeded onto an Immobilon P filter. and cocultured with a segment of chicken E4 hindgut on the chorioallantoic membrane. After 7 days ot" coculture. we found that quail vagal neural crest cells. cultured for 1 day in vitro. were still c::tpable of forming enteric ganglia (n = 6). The enteric ganglia, however. were some­what smaller and fewer than cocultures with vagal neural anlage. Staining with the HNK-l ::tntibody revealed en~ teric ganglia (HNK·1 mode 2). although there was still some HNK·l-immunoreactivity present within the enter­ic mesenchyme (Fig. 3A). A minimum of ten thousand vagal neural crest cells per filter. equivalent to the out­growth of 5-10 neural anlagen. was necessary to obtain colonization of the gut. The same 'lumber of cultured neural crest cells from trunk colonized the gut at the

103

Fig. 2A-C. Pamain sections of 7-day cocultures of E2 quailncural anlage and E4 chicken hindgut. A \.ragal neural anlage: stalning with the HNK·l antIbody shows mode 2. visualizing the myenteric (.1") and the submucous (S) ganglia. Remak's ganglion (R) is also stained. :.:40 B Trunk neural anlage: staining with the HNK·j antibody shows HNK·j mode j.immunoreactivity. visualizing a layer of immunorcactive cells within the enteric mesenchyme at the Slte of the submucous ple:o;us.es. Within this layer mc!unocytes are present (arrows). x 40 C Detail of B showing melanocytes (ar­row). x 63

proper sites but did not give rise to enteric ganglia (n "'" 3). Instead we observed me1anocytes and more-intense HNK-I mode 1 immunoreactivity (Fig. 3 B). Thus. the difference in differentiation seen earlier between vagal and trunk neural anlage was based on an intrinsic differ­ence between vagal :lnd trunk neural crest cells. The prolonged presence of the neural tube and notochord after 24 h had no !!ffect on neural crest cell differentia­tion.

104

Intrinsic dijJerences he{\\"een L'aga! and trunk neural crest cells can he illustrared in t;itro

Som!! characteristics of vagal and trunk neural crest cells were studied in vitro. First. vagal and trunk neural crest cells were compared aft~r 1 day of in vitro culture, at the tim~ vagal neural creSt cdls were capable of forming enteric ganglia in the colonization assay. Vagal neural crest cell cultures consisted of about one to two thousand small t1attencd stellate cells. After fixation and staining with the HNK-l antibody. we observed 40%-50% HNK·1 positive cells tbased on the counts of six cuI· tures; Fig. 4A). Trunk n~ural cr..::st cell cultures consisted of one to two thousand cells. which were small and round to stellate. These cultures consisted of a higher percentage (i5%-80%.) of HNK·l positiv..:: cells (bas..::d on the counts of five cultur~s: Fig. -1-8). Occasionally, small clusters of epithelial-like cells were observed in both vagal and trunk neural creSt cell cultures (Fig. 4A. arrow). -These could have been remnants of ectoderm or the neural tube. We then studied vagal and trunk neural crest cells upon further culturing. During the first week of culture. vagal neural crest cultures showed mod­erate proliferation (determined by the increase in cell number: data not shown). Th!! percentage of HNK-l· positive cells rapidly declined during the first -1- days of culture. from 40'%-50% on day 1 to 5%-10'% on day 4. and remained low during the next weck of culture. Trunk neural crest cells proliferated rapidly. showing a vast increase in cell number (data not shown). During the first week of culture. the percentage of small stellate HNK-l-positive cells remained high (85%-90%). After the first week it gradually declined. to about '::'0% aft~r 3 weeks. In Fig. 5 the changes in the percentage of HNK­I-positive cells during culture of vagal and trunk neural crest cells are shown.

The differences in morphology and H"NK-l expres­sion between vagal and trunk neural creSt cells after 1 day of culture in vitro illustrate an intrinsic difference between these two cell populations. Th!! decline in HNK-1 expression of vagal neural crest cells upon prolonged in vitro culturing indicates that vagal neural crest cell prop!!rties change during in vitro culture.

In uirro changes 0/ cagal neural crest cell properties affect their differentiation in the hindgut

To investigate wh~ther the changes ot' vagal neural crest cell properties observed during in vitro culture in-11uenced their differentiation behaviour. we tested vagal neural crest cells. cultured for 4 days in vitro, in our colonization assay (n "'" 5). We found that vagal neural crest cells were still able to colonize the gut at the proper sites. but they were no longer able to differentiate into enteric neurons. Instead. melanocytes developed in the gut (Fig.6A). Trunk neural crest cells cultured for 4 days still gave rise to melanocytes (n=3: Fig. 6B).

These data further support the hypothesis that there is an intrinsic difference between vagal and trunk neural crest cells just after migration from the neural anlage.

Fig. 3A. B. P,lr::lt'tin sections of 7-d,ly cocultures of qll,lil vagal nellral .:rest cells. culwrcd for 1 day in vitro. and E'" chicken hmd{!llt A Vagal nellral crest celts: the H:;..iK-l antibody visllali:>:es myenteric (,\.1) J.nd submllcous (S) ple.~l!Ses. There is still .;ome HNK-I mode i-immunoreactivltv visible. x40 B Trllnk neun.tI cre~! cells: the HNK-I antibodv ViSllalizes mode l-immllnoreaCllvitv. and no enteric neurOns are prese~t. x"'O .-

Fi1/:. 4A. B. QU:lil neural crest cells cultured for 1 duy in vitro. A VagJ.l nellral cre~t cells: slaining with H'N'K-l reveals about 40%-50% of HNK-l-immunoreactive cells. which are of a stellate. t1attcned morphology. The smull cillster of epithelial-like cells present in the culture (arrow) is probably a neural tube remnant. B Trunk neural creSt cells: Staining with HNK-l reveals a high percentage (75%-80%) of HNK-l-immunoreactive cells of a small. rollnd to stellate morphology. x 40

100 rl ------------,

'" 90 I ~ 80~... h,

70' \ / \

.~soV "-&. 50 ' ..... /' .....

"-"­

"-"-

" ~~~~ " 20

10r OL----~5c----,~0--C-~,~5----2~0C---~25

days in culture Fil!. 5. Changes in the percentage of HNK-I-cxpressing cells in neural crest cell cultures. The solid line represents vagal neural crest cells. the dashed line represents trunk neural creSt cells

The specific characteristics of vagal neural crest cells which enable them to form enteric ganglia in the.postum­bilical gut are lost upon prolonged culturing in vitro. They retain their colonization properties. but they now differentiate similarly to trunk neural crest cells.

Discussion

Differentialion of vagal and lrunk neural crest in postumbifical gut

Neural crest from both vagal and trunk levels could colonize the hindgut. but under our experimental condi­tions. formation of enteric ganglia was confined to vagal neural crest cells. Trunk neural crest cells migrated and homed to the correct sites but they mainly differentiated into melanocytes. i'Jeural differentiation sometimes oc­curred. but enteric ganglia did not develop. This implies that neural crest cells from various axial levels are able to recognize migration and homing signals in the enteric microenvironment, but that cells from the vagal neural crest are better capable of responding correctly to differ­entiation signals.

Le Douarin and Teillet [1SJ introduced the chorioal­lantoic membrane coculture method as an assay for neu­ral crest cell colonization of the hindgut and in this way established pluripotentiality of the neural crest. They showed that trunk neural crest cells can give rise to enter" ic ganglia. and this method was subs~quently used in a number of studies [23. 32, 34]. All those studies de·

105

F;~. 6A. B. Paraffin sections of I.day .::ocultures of E:: quail no;:ural crest eells. eultured for..l. days in vitro. und E4 chicken hindgm . . -\. Vagal neural crest cells: :Vinny melanocyte':' are in the submucosa (arrows) and there are als some melunocyt<:s ;.It the site of the myenteric plcxus (arron-he'ads). :..; 40 B Trunk neural crest o:lls; there arc many melanocytes in the submucosa lam)wh(fJds). x..l.O

scribed enteric ganglia in the hindgut. but the number of ganglia varied. possibly in relation to the experimental system used. A neur to nomlal number of enteric ganglia was found. when E4.5 or E5 chicken hindgut was cocul~ tured with E2 quail neural anlage. which was introduced into a slit made in the hindgut and subsequently cultured for 12-24 h in vitro before placing it on the chorioal1an~ toic membrane [18. 32. 34]. In direct cocultures of E4 hindgut and E2 neural anlage. without making a slit and omitting the in vitro culwre. enteric neurons were extremely sparse with many sections not containing neu­rons [23}. Smith et al. [32] attributed this difference be~ tween vagal and trunk neural crest segments to a quanti~ tative difference between these two neural crest regions. the vagal neural crest being larger than trunk neural crest. We found. however. that an amount of vagal neu­ral crest equivalent to the length of two somites is al~ ready sufficient for the formution of a normal enteric nervous system in this coculture system (unpublished results). All these coculture studies agree on the presence of me1anocytes in the hindgut after colonization with trunk neural crest. which is in line with our results. The capacity of trunk neural crest cells to give rise to !:!nteric neurons was also tested in heterotopic quaikhick chi~ mera in which the adrenomedullo.ry crest was trans~ planted to the vagal level. In the preumbilical gut. trunk neural crest gave rise to 0. norm::tl enteric nervous system. whereas in the postumbilical gut the quail cells found in the gut wall were exclusively pigment cells and never participated in go.ngli.on formation [is]. Additional evi~ dence for a specific intero.ction betwe!:!n vagal neural crest cells and the gut comes from the same study. in which quail vagal neural crest was transplanted to the adrenomedullary region of a chicken embryo [18J. The vagal neural crest behaved as adrenomedullary crest ex~ eept that in the gut there were quail cells. which had differentiated into enteric neurons. In normal develop~ ment trunk neural crest cells never penetrate the splanchnic mesoderm. indicating that at this level no preferential route leads the cells to the intestine [is]. These data indicate that although trunk neural crest cells can give rise to enteric neurons under certain experimen~ tal conditions, they differ from vagal neural crest cells

106

regarding their differentiation into me1anocytes in the postumbilical gut.

The role 0/ the enteric microenvironment in the/ormation 0/ the enteric nert·ous system

In this study we found that the HNK-l-immunoreactivi~ ty within the enteric mesenchyme of ::meural gut disap· peared only when colonization by neural crest cells was followed by the formation of'enteric ganglia. After colo­nization with trunk neural crest the HNK-l~immunore~ activity persisted within the enteric mesenchyme. Thus the switch from mode 1 to mode 2 Hi-JK-l-immunoreac· tivity correlates with the differentiation of neural crest cells into enteric neurons and the subsequent formation of enteric ganglia. and not with mere colonization. These results indicate that there is an interaction between neu~ ral crest cells and the enteric microenvironment.

A number of studies describe an influence of the en~ teric microenvironment of neural crest cell differentia· tion. Dulac and Le Douarin [5] found that the Schwann cell marker SMP on enteric glial cells is down· regulated by the influence of the gut wall. Pomeranz et 0.1. [28] suggested that an interaction between laminin in the en· teric microenvironment and a laminin binding protein on neural crest cells plays a role in the homing of neural crest cells in the gut und the subsequent development of enteric ganglia. Galliot et al. [7] described expression of the Hox~l.4 gene in fetal mouse gut mesenchyme dur· ing normal embryonic development. Wolgemuth et a1. [37] made transgenic mice overexpressing the Hox~l.4 gene and found that this resulted in abnormal gut devel~ opmcnt. It was found that overexpression of Ho:(A.4 resulted in a nonfunctional enteric nervous system. be­cause the neural crest cells differentiated into catechola· minergic neurons. displaying the Ultrastructure of pe~ ripheruL not enteric. nerves [8].

Considering these results. it is possible that the enteric microenvironment in the post umbilical gut produces a signal necessary for the differentiation of neural crest cells into .::nteric neurons. Vagal neuro.l crest cells would then respond to this signal and differentiate into enteric

neurons. whereas trunk neural crest cells are less able to recognize this signal or react to it in 0. different way.

Intrinsic difference hetween L'agal and trunk neural crest cells iliustra£ed in vitro

We developed 0. new coculture syStem. which uses in vitro cultured neural crest cells. and demonstrated that the prolonged presence of the neural tube and notochord after 24 h did not influence the difference in neural ver­sus melanocyte differentiation between vagal and trunk neural anlage in the hindgut. In the absence of the neural tube. a minimum of ten thousand vagal neural cr!!st cells. cultured for 1 day, was n!!eded to achieve colonization and to fOfm !!nterlc ganglia. The enteric ganglia. how­ever. were somewhat smaller and fewer than with vagal neural anlage. and there was still som!! HNK-I mode I-immunoreactivity present. During the fonnation of en­teric ganglia. HNK-l mode 1 immunoreactivity gradual­ly disappeared. In chorioallantoic m!!mbrane cocultures using vagal neural anlage. HNK-j mode I had disap­peared by the 7th culture day (personal observation). Using cultured neural crest cells could cause a delayed or incomplete disappearance of HNK-l mode 1 immu­noreactivity. This could be related to the number of n!!u­ral crest cells used. or to a change in vagal neural crest cell properties even during the first day of in vitro cul­ture.

We attempted to correlate the observed difference in dilTerentiation in vivo with an intrinsic difference be­tvleen vagal and trunk neural crest cells in vitro. We found that 40%-50% of the vagal neural crest cells showed HNK-l-immunoreactivity after 1 day of in vitro culture. whereas trunk neural crest consisted of 75%-80% HNK-l-immunoreactive cells. There are a number of studies on the expression of the HNK-I epitope on trunk neural crest cells in vitro [22. 29, 31. 36]. Our result of 75%-80% HNK-l-positive trunk neural crest cells after 1 day of culturing agrees with three other studies reponing the percentage of HNK-l-positive cells after 1 day of culture [22. 29, 35]. In these studies there are no data on the percentage of HNK-l~immunoreac­tive cells in vagal neural crest cell cultures. It is not dear whether the observed difference in HNK-I expres­sion betvleen vagal and trunk neural crest cells in vitro reflects a similar difference in vivo. Up to now the HNK-1 antibody has been the only marker for early migrating neural crest cells in chicken embryos and therefore it was not possible to detect HNK~l-negative neural crest cells in vivo. With the recently developed lineage tracer techniques. such as DiI and retroviral markers (6. 21, 27J. it will be possible to study neural crest cell develop­ment in vivo independent of HNK-l-immunoreactivity.

On prolonged culture. the percentage of Hi'JK"l~im­munoreactive vagal neural crest cells rapidly declined to 5% during the next 3 days in vitro, indicating a change in characteristics of these cultured cells. Vincent and Thiery [36] showed that a number of neural crest derivatives lose their HNK-l epitope and that only neu­ral crest derivatives related to the peripheral nervous

syStem retain this epitope, The observed drop in percent" age of HNK-l-immunoreactive cells is concomitant with a loss of capability to differentiate into enteric neurons. Whereas after 1 day of culture vagal neural crest cells showed normal colonization and differentiation proper­ties. after 4 days of culture they lost the ability to form enteric neurons and instead ditTerentiated into melano­cytes.

In conclusion. we sunnise that the fonnation of the enteric nervous system in the postumbilical gut entails at least two phases. First, neural crest cells migrate through the gut and home to the sites of the mycnteric and the submucous plexus. This colonization phase is not specific for vagal neural crest cells: crest cells from other axial segments are also able to colonize the gut. During ~he second phase. ncural creSt cells differentiate into enteric neuron.:; and fonn enteric ganglia. for which the enteric microenvironment probably provides a sig­nal. Correct differentiation into ent!!ric neurons depends on intrinsic prop!!rties of vagal neural crest cells. In our e.'(periment.:'11 system, most trunk neural crest cells either cannot recognize the signal or react to it differen.tly and differentiate into melanocyteS. The switch from mode 1 to mode 2 HNK-l immu'noreactivity oCCurs upon for­mation of enteric ganglia, during the second phase.

Acknowkdg'·!1U'nlS. The authors thank Dr. M.P. Mulder for critical reading of the manuscript. Prof J.C. Molenaar lor his continl.1ol.1s support. and Mr. T. de Vries-Lentsch for photography. This study was supported by a research grant of the ~ethcrkmds Digc:>tive Di:)<:a:)<:s Foundation (grant .q.WS 89-21) und the Sophia Founda, tion For Medical Rcsearch (f':mnt * 105).

References

Abo T, Balch CM (1981) A differentiation antisen of human NK and K celts identified by a monoclonal antibody (HNK-l). J [mml.1noI127: 102..1-1029

2. Altan 11. ;.Jewgreen OF (1980) The origin and differentiation of enteric neurons of the intestine of the towl embryo. Am J Ann! 157: 137-154

3. Bockman DE. Kirby :VIL (1984) Dependence of thymus devel­opmcnt on derivatives ot' the neural crest. Science ::::3 ;498-500

4. Bronner·Fruser M (1990) Experimental analysis of the migra­tion and celt lineage of avian neural crest celts. Cleft Palate J 27: 110-120

5. Dulac C. Le Douarin I':M (1991) Phenotyp'lc plasticity ot' Schwann cells and enteric glial ee!ls in response to the microen­vironment. Proc Natl AC:1d Sci USA 88:6358-6362

6. Fraser SE. Bronner-Fraser M (1991) Yiigrating neural crest cells in the trunk of the avian embryo are multi potent. Develop­ment 112:913-920

7. Galliot B. Dolk P. Vigneron M. Featherstone MS. Baron A. Duboule D (1989) The mouse Hox-l.-1- gene: primary structure. evidence tor promoter activny :md expression during develop­ment. Development 107:343-359

S. Gershon MD. Tennyson VM (1991) Microenvironmenta! t·ac· tors in the normal and abnormal development ot' the enteric nervous system. !n: The Morphogenesis of Down Syndrome. Wiley-Liss Inc. New York. pp :::57-:::76

9. Hall BK. H6madius S (1988) The neural cn:st. OKford Univer­sitv Pre~$. Oxford

10. Hnmburger V. Hamilton HL (1951) A Senes of normal stages in the development of the chick embryo. J Morphol 88:':'9-67

107

ii. Hollman S, Edelman GM {\987) A protco£:lycan with H~K- I antigcnic determinants is a nC'.lron-aSsocialcd ligand lor cyto­tactin. Proc 1"atl Acad Sci USA 84::523-:5:7

I::'. K:J.lcheim C. Bardo! YA. Th(X:nen H, Le Douarin;-.lM 0937) In vivo effect of brain-derived neurotrophic factor on the ~ur­vival of developing dorsal root gunglion cdl$. E:vr BO J 6: ::87 1-::873

13. Kirby Ml (1939) Plastidty and predetermination of mcsenCe­phalic and trunk neumi crest transplanted into the region 01 the cardiac neural crest. Dev Bioi 134:402-..\1::

14. Kirby :vIL. Gale TF. Stewart DE (1933) Neural crest cells con­tribute to normal aorticopulmonary septation. Science :20:1059-1061

15. Kruse J. :vIailhammer R. Wcrnecke H. Faissner A. Sommer l. Goddis C. Schachner M 11934) :--.Ieurul Cell :.ldhesion mole­cules and myelin-assocw.ted glycoprotein share a common C~lr' bohydratc moiety recognized by monoclonal antibodks L: und H>l"K-1. Nature311:153-155

16. Le Dou:J.rin NM (198:) Thc neural crest. C"lmbrid~e Univ. P~ess, C:J.mbridge ~

17. L~ Douarin NM, Teillet M (1973) The migrJ.tion of neural crest cells to the w:J.ll of the digestive truCt in avian embryo J Embryo! Exp MorphoI30:31-4S

IS. Le Douarin NM. Teilkt M (1974) Experimental analysis of the migration and dilTerentiation ot' neuroblasts of the .luto­nomic nervous system :J.nd of neuroectOdermal mesenchymal derivatives. using a biological cell marking technique. Dev Bioi 41:162-184

19. Le Li<i:vre CS. Le Dou:J.rin NY'! (\975) Me:;enchymal deriv:J.tives ot' the neural crest: analysis of chimaeric qU:J.il :J.nd chick em­bryos. J Embryol Exp Morphol34: 125-154

:0. LlJiderTM. Peters-van derS:J.nden MJH. Molenaar JC. Tibboel D, van der Kump AWM. Meijers C (199:) Ch:J.ract~rization of HXK-1 antig~ns during the form:J.tion of the ;lvian enteric nervous system. Development 115:561-572

21. Lumsden A, Sprawson N, Gr:J.ham A (1991) Segmental origin and migration ot' neur:J.1 crest cells in the hindbr:J.in region of the chick embryo. Development 113: 1231-1291

22 Maxwell GD. Forbes ME. Christie DS (1988) Analysis ot' the development ot' cellular subsets present in the neural crest using ccll sorting and cell culture. :--.I"curonl :557-563

23. :--.I"ewgreen DF.Jahnke!, Allan IJ, Gibbins IL (1980) Differenti­ation ot' sympathetic and enteric neurons of the fowl embryo in gr:J.lb to the ehorio-allantoic mcmbrane. Cell Tis~ue Rt."S 208: \-19

24. Noden OM (1978) Interactions directing the migrution and cy­todifferentiation ot' avian ncurJ.1 crest Cells. In: G:J.rrod D. The Specificity of Embryological Interactions. Chapman and Hall. London. pp 4-49

108

25. Noden OM (1991) Vembr.lte cr:J.niof:J.ci:.ll development: The rebtion betwet:n ontogenetic process und morphologlc:J.l out­come. Brain Behav Evo138;190-225

:6. PomeranZ HD. GershOn:Vro (1990) Colonization or the aviun hindgut by cells derived from the sacral neural crest. Dev Bioi 137:378-39..1-

27. Pomeranz HD. Rothman TP. Gershon YlD 11991) Coloniza" tion of the post-umbilical bowel by cells derived l'rom s:"lcf:.ll neur:J.1 crest: dircct tr;lcing ot' cell migr:J.tion using an interC""J.lat­ing probe :.lnd a replication-deliden-t retrovirus~ Development 111:647-655

28. Pomemnz HD. Sherm:J.n DL Smalheiser NR, Tcnnyson VM. Gershon MD (1991) hpression ofa neurally related bminin binding protein by neuml crest-derived e~lIs that colonizc the gut: relationship to the formation of ~nteric £;anglia. J Comp >Ieurol 3 I 3: 625-6J.:

::9. Sanders EJ, Cheung E (1938) EITe<:t:> ot' HNK·I monoclonal antibodv on the sub,;tratum :J.!tachment and sunivJ.1 of neural crest :.lnC! sclerotome cells in culture. J Ceil Sci 90: 115-12:

30. Serbedzija GN. Burgun S, F~aser SE. Bronner-Fraser YI (1991) Vital dye labelling demonstrateS :J. sacf:J.lneural crest contribl1-tion to the enteric nervous syStem 01' chick and mouse embryos. Development III : 857-367

31. Sieber-Blum M, Kumar SR. Riley DA (1988) In vitro ditTerenti­ation of qU:J.ll neural crest cells into sensory-like neurobi:J.sts. Dev Brain Res 39: 69-$3

3:. Smith J. Coch:J.rd P. Le Douarin NM (1977) Development Ot choline acctyltro,nsfcrase and cholinesterase uctivities in enteric g:J.nglia derived Crom presumptivc adrenergic and cholin~rgic levels of neural crest. Cdl DiIT 6: 199-:16

33. Teillet M. K..11ch-.-im C. Le Douarin ;"\1M (1987) Fonn..1.tion of the dorsal root ganglia in the avi:ln embryo: Segmenml origin :.lnd migratory beh:J.v\Our of neur:J.l crest progenitor cells. Dev Bioi 120:329-3.:17

34, Tcillct MA. Cochard p, Le Douarin ~M (1973) Rdative roles ot' the mesenchym:tl ti$.Sues and ot' the complex neurul tube­notochord on the expression 01 :J.drcnergic membolism in neural cre:;t cells. Zoon 6: 115-122

35. Vincent M. Duband J, Thiery J (1983) A cell surface determi­nant expressed early on migrating :J.vian neural cre:;t cei1s. Dev Brain R~s 9:235-:38

36. VinCent M, Thiery J (1984) A cell surf:tce m:trker tor nelJfal crest and placodal cells: further evolution in peripheral and central nervous system, Dev Bioi 103:.:168~Sl

37. Wolgemuth DJ. Behringer RR. Ylostoller MP. Brinster RL, Palmiter RD (1989) Tr:J.n$genie mice ovcre:-::pressing the mouse homeobox-contuining gene Ho:-::"I.4 exhibit abnormal gut de­velopment. >l"atl1rc 337:..t64-467

38. Yntem:J. CL. Hammond WS 11954) The origin of intrinsic g:J.n­gl;a of trunk Viscera I'rom v:.t,g<ll neural crest in the chick em­bryo. J Comp NeurollOl :515-541

DevtlopmeM 11:5. 561-57.: r 199:) Printec:! in Greut Britam © The Company ot' 8,olo~I~(s Limit~d 1<l92

Characterization of HNK-1 antigens during the formation of the avian

enteric nervous system

THEO M. LUIDERI. .. , MARlO J. H. PETERSNA:"J DER SANDEN l . JAN C. MOLENAAR l . DICK

TIBBOEL1. ARTHUR W.:VI. VAl\' DER KAMp':: and CAREL MEIJERS l

'Department oj P~dialrlc Surgery. Erasmus Universir,v. Sophia Chrld'en's Ho~pl/(l/. ROlrerdam. till: S~lhcrfands :.'fedica! Gen~lics Cenrer ROlladam-Lelden. D~partmt'11I oj Cd! Biology and Gmau:s. £rasnws Cm'·crSily. ROlladum. rhe Saher/ands

·Corre~pondins al.ltnor. at: Department of Cdl B,oloSY :lnd Genetics. Erasmus Unive~ity. PO Box 1738. 3i'J()() DR Rotterct~m. the >fetnerlands

Summary

During vertebrate embryogenesis. interaction between neural crest cells and the enteric mesenchyme gives rise to the development of the enteric nervous system. In birds. monoclonal antibody HNK·l is a marker for neural crest cells from the entire rostrocaudal axis. In this study. we aimed to characterize the HNK~1 carrying cells and antigen(s) during the formation of the enteric nervOus system in the hindgut. Immunohistological findings showed that HNK.l~positive mesenchymal cells are present in the gut prior to neural crest cell colonization. After neural crest cell colonization this cell type cannot be visualized anymore with the HN'K-l antibody. We characterized the HNK·l antigens that arc present before and after neural crest cell colonization of the hindgut. Immunoblot analysis of plasma membranes

Introduction

Glycosylation and sulphatation are important post­translational modifications of proteins serving essential roles in embryonic development (Sorkin et a1.. 1984: Feizi. 1985; Imamura and Mitsui. 1987). Thorpe and coworkers (1988) suggested that carbohydrate struc­tureS play crucial roles in intercellular interactions. All proteinaceous cell ::tdhesion molecules described so far are glycoproteins. but the function of the carbohydrate moieties has onlv been elucidated in a few cases (Hoffman and Edelman. 1983; Sadoul et a1.. 1983). One particular carbohydrate moiety, a complex sulphate-3~ glucuronyl residue. is recognized by monoclonal anti­body HNK·l. Chou and coworkers (1985. 1986) characterized the molecular structure of the HNK-l epitope both in a glycolipid and in a tetrasaccharide which had been isolated from human peripheral nerves. It is likely that the carbohydrate moiety on glyco­proteins is similar if not identical to the cpitope on the glycolipid and the tetrasaccharide (Shashoua et al.. 1986: Burger et a1.. 1990). The HNK~1 epitope is

from embryonic hindgut revealed a wide array of Hl'\K­l·carrying glycoproteins. We found that two a.~K-l antigens are present in E4 hindgut prior to neural crest cell colonization and that the expression of these antigens disappears after neural crest colonization. These two membrane glycoproteins. G~2 and G • .w. have relative molecular masses of 42.000 and 44.000. respectively. and they both have isoelectric points of 5.5 under reducing conditions. We suggest that these HNK·l antigens and the HNK·l-positive mescnchymal cells ha\'c some role in the formation of the enteric nervous system.

Key words: HNK·l. nCltrai crest. cd! adhesion molecule. enteric nervous system.

present on a series of molecules involved in cell adhesion. substratum adhesion and extracellular matrix interactions (Kruse et a1.. 1984: Faissner. 1987: Pesheva et 31.. 1987; Hoffman and Edelman. 1987). Keilhauer and coworkers (1985) and KUnemund and coworkers (1988) demonstrated that the HNK-l epitope itself is involved in neuron~neuron and glial~glial cell interac­tions in in vitro adhesion assays. Bron~er-Fr:.'lser (1987) showed that injection of the HNK~1 antibody later:.'ll to the mesencephalic neural crest of chicken embryos even perturbs neural crest migration in vivo.

In chicken embryos. HNK~l antigens are present at very early stages of development (Canning and Stem. 19S5). Using immunoablo.tion. Stern and Canning (1990) found- that ffu'K·l~positive cells playa key role in gastrulation. During neurulation. HNK·l visualizes premigratory and migrating neural crest cells. It is accepted that the HNK-l antibody can be used as a marker for avian neural crest cells (Tucker et a1.. 1984; Vincent et al.. 1983: Tucker et a1.. 1986: Newgreen et al. 1990). although structures not derived from the neural crest can also be HNK-l positive (Serbedzija et

109

T. :vf. Luider and others

al.. 1991). At late, stages of development the HNK-l monoclonal antibodv visualizes rhombomeres three and five (Kuratani -et a1.. 1991). and the central. peripheral and enteric nervous systems. The HNK-l epitOpe has also been found in the nervous systems of other vertebrute and invertebrate species (Schwarting et al.. 1987: Mikol et al.. 1988: Tucker et ai., 1988: Dennis et a1.. 1988: Poltorak et al.. 1989: ;.Jordlander. 1989: Metcalfe N o.l.. 1990). Interspecies differences in HNK-l expression in the brain have been reponed (O'Shannessy et aI., 1985: Holley ::tnd Yu, 1987), In addition, the expression of H:-';K-l antigens in various structures changes during development (Mikol et at.. 1988: Newgreen et at. 1990).

In the chicken embryo, the neurons and supportive cells of the enteric nervous system develop from rhombencephalic (vagal) neural crest cells. which emerge at the level of so mites 1-7. and sacral neural crest - cells. which emerge poswrior to somite 28 (LeDouarin and Teiller, 1973). Using the HNK-l amibodv to visualize neural crest-derived cells in the wall o(the gut. Epstein and coworkers (1991) found that crest-derived enteric precursors form a cellular network when they reach the primordial gut distal to the lung buds. The authors suggest that this network is probably a precursor to the ganglionic network in the adult proximal gut. The behaviour of the crest-derived cells. which underlies the formation of this network. is different from the behaviour of crest-derived popu­lations forming autonomic and sensory ganglia. The microenvironment of the gut may be a major com­ponent in producing this different behaviour. There is no knowledge concerning the molecular natLlre of the HNK-l antigens during -the formatiOn of the enteric nervous system,

In this studv, we aimed to characterize the HNK-l­carrying cells ~nd antigc:!n(s) during the formation of the enteric nervous system in the hindgut. We therefore invesrigatc:!d explants of the hindgut at various stages of development with immunohistochemical and biochemi­cal techniques.

Materials and methods

Embryos Fertilized egg5 of Gallus gallus domcsficus were obtained from 0. local supplier o.nd incubated in 0. forced draugn.t incubator at 37'C and 80% humidity. Embryos were staged according to tn.e number of incubational days (E=day of development) or to the table of Hamburger and Hamilton ( 1951).

Exp{aruarion of embryonic hindgut Gut segments were iSOlated between the ceca and the cloaca. We us';d E4 till E14 embryos. The mesentery was removed.

Chorioallantoic membrane cultures Se"ments of 1 mm hindgut were grafted OntO the chorioallan­:oi~ membrane as described prev-jously (Mdjers et al.. 1987). E4 hindgut was h:,ll-ve$t<:d after seven days and Ei hindgut after foLir days' culture. The chorioallanwic membranes and

110

blood vessels were removed and the dis.sected I!rafts were homogenized or prepared for immunOhistochemistry. For coculture experiments. ,he vagal neural primordia adjacent to the t'irst s~ven somites of stage 10 embrYos were dissected and cocultured with E4 and E7 hindgut for Seven and four davs. re~pectively. - .

Immunohisrochemisrry After dissection. gut segments were rinsed in phosphate­buffered saline (PBS) and fixed overnight in 4% paraformal­dehyde-PBS. dehydrated and embedded in paraffin (Fluka, Switzerland). SectiOns were made at 5-7 .wn. Alternatively. sel!mems were embedded in Tissue Tek II embeddin~ compound (Milcs, :-.:aper .... ille. IL) and snap-frozen in liquid nitrogen-cooled isopentane. Sections wt::re made at lO .um. Sec,ion$ were incubated with the primary antibody in a moist incubation chamber at room temperaturc for one hour. For immunofluort::scencl;:, rabbit anti-mouse FITC-conjugated F(ab)~ fragments of immunoglobulins (Dako. Denmark) were used as a second step antibody (diluted 1:20). For immuno­peroxidase staining. rabbit anti-mouse peroxidase.conjugated immunoglobulins (Dako. Denmark) were used as second step antibodie$ (diluted l:lOO). Endog<:nous peroxidase$ were inhibited by a 20 minute incubation in methanoVhydrogen peroxide (99:1/ Vjv) solution. Peroxidase w:\s visu::tlized with 0.1% 3.3'diaminobenzidine.HCI (Sen'a, FRG) with 0.01% hydrogen peroxide. Sections were counterstained with hema­tOxvlin for one minute. PBS with 0.1 % Tween-::'O W:l.S used for ::til ·rinsing. Sections were evaluated using a Leitz Fluorplan microscope, or with a Biorad Confocal Laser Scanning Microscope mounted on u :-:ikon tluorescence microscope.

Amisera The HNK-1 hybridoma was purcn.aSl::d from the American Tissue Type Culture Collection (TIB200) (Abo and Balch. 1981). Cells were grown in RPMI medium (Life Technologies. Breda. the :-.Ietherlandsl supplemented with 10% fetal enll' serum (Sunbio, Uden. the Netherlands). penicillin a.75 mgtml. streptomycin 125 mg/ml and glutamine 2.92 mgt'ml (Life Technologies. Breda. the :\etherlands). Conditioned media were harvested after three davs' culture. Monoclonal antibody E/CS was purchased fro'm thr.: Developmental Studies Hvbridoma Bank. mAb E;'CS is directed '::Heainst NAPA-i3: a neurofilament-associated glycoprotein.~ The $upernatant of the hybridoma culture was used undiluted (Cimem and WestOn. 1982: Ciment et 0.1.. L986).

Homogenisarion of tissues for plasma membrane analysis Plasma membranes were obt:lined according to 0. modification of the prowco! of :-Vfaeda and coworkers (l983). In brief. ,issues were collected in TSE buffer (10 mM Tris/HCl pH 8.0. 0.25 M sucrose. 1 mM EDTA) at 4°C. homogenized in an Omni Mixer Homogenizer (Connecticut, liSA-) on ice for 1 minute (level lO). ~The suspension was centrifuged for 5 minutes at :;:oaa revs/minute in an Heraeus centrifuge. The supernatant was layered W:l solution containing 41% sucrose. to mM Tris/Het pH S.D. t mM EDTA. and centrifuged in a Beckman ultracentrifuge for 1 hour at 24.000 revs/minute in 0.

SW"-8 rotor. The int;rphase containing the plasma mem­branes was collected and diluted with TSE buffer and centrifw2ed again for 1 hour at 24.000 revs/minute in a SW::'8 rotor. The peliet was re$uspended in TSE buffer and stored at -7a~c. Protein content wa$ determined with the BCA assay (Pierce USA). We purified 105 .ug. plasma membrane protein from 35a explo.nts of E4 gut (total protein content per explant

40 .ug prot(!in): 69 (!;>;pl.:lnts of E7 gLl.t contained 1.4 mg plasma membrane protein: 10 e;>;plants of E14 gut contained 3.0 mg plasma membrane: -10 cultLl.res of E-t. hindgLl.t contained l,2 mg plasma membrane proteins: ::0 cultures of E7 hindgut conto-wed 0.5-1 mg plasma membranes.

Two-dimensional gel etecrrophoresis Gels with a length of 6.5 cm were prepared in gbss tLl.bes with a diameter of 2 mm according w the maoLtfacturer's description (Biorad, California). In brief, 0.25 ml Biolyre 3/10 ampholyte, 0.25 ml Biolyte 5i7 ampholyte, 2 ml 10% Triton X-lOa. 5.5 g urea analytical grade (Merck, FRG), 1.33 ml acrylamide solution (28.3% :lcrylamide :lnd L62% piperazine di-:lcrylamide (Biorad, California), 1.97 ml distilled water. to .ul 10% ammonium persulfate, 10 .(11 :-:.:-r S' ,N' -tetra-metbv­lethylenedi:lmine (Btorad, California)) wert: mixed and 3.11owed to polymerize in glass tubes at 37°C. Pre-elcctrophor­(!sis was perfonned at 200, 300, -100 V for 10, 20 and 20 minutes. respectivdy.

For isoelectric focLtsing. samples were diluted with an equal volume of sample buifer (2% SDS. 10% glycerol. 62.S mM TrisiHCI pH 6.8, bromophenol blue. 0.1% dithiotreitol (CalbiOchem, California), boiled for 3 minutes and chilled on ice. 50S was added to the protein sumple to facilito-w solubilization. Samples w<:re then diluted with an eqLtal volume of lysis bLtffer (9.5 M urea. 2% Triton X-lOO. 0,1% dithiotreiwl. 1.6% Biolyte 517 ampholyte. 0.4"/0 Biolyte 3/10 ampholyte in distilled water). The SDS was removed from the proteins by the Triton X-100 micelles. SD.mplcs of 100 .ul containing 50 ,ug of proto::in were loaded under D.n ovo::rlav solution and subjected to electrophoresis for 3,5 hoLtfS CIt 60b v.

For so;:paration in the second dimcnsion. the gels were g<:ntly removed from the tubes and equilibrato;:d for :l.pproxi­matdy 90 minut<!s in sample bLtffer until the pH indicator in tho:: D.cid part of the gel became blue. The go::! tubes w<:re 10:ldcd directly on 0. 2.25 mm thick 7.5% 50S-polyacrylamide minigel (Biorad, California) and subjected to electrophoresis for 15 minLttes at 100 V and then at 200 V Ll.ntil the bromophenol blue reached the bottom of the gel. C.:lrbamy­lyte creatine phosphokina:>e (Pharmacia, Sweden) was used:LS a stand:lrd for isoelectric focusing :lnd pr~stained protein molecul::tr weight standards (14.3 - 200 x 1O~ ,\1,) !Bethesd:l Rese:lrch Laboratories, :vID) were used in the second dimension.

HNK·j antigens and enteric neural crest

HNK-J immunoblotling Proteins were transferred from the gd onw a 0.-15 ,tim nitrocellulose SSBAS5 membrane (Schleicher and Schuell. FRO) in a Biorad blot :lpp:lratus at 200 mA and 100 V for 90 minLttcs in a blotting bLtffer containing 20% (v/v) methanoV50 mM Tris/Glycine p~H S.O. Blots were blocked by overnight incubation in 2% bovine serum albumin (Sigma. Fraction V. St. Louis) in PBS-Tween-:'O (O.l'~~) at -1~C and 20 minLtleS incubation in L % normal gO:l.t serum (Amersham Inter­nationD.1 Pic, UK) in PBS-Tween-20 at room lemperatLtre, Subsequentlv, the blots were incub:lted with :l twentv times diluted sup~rnatant of HNK-1 hybridoma cLtlture 't'or -15 minutes :n room lemper:ltLI.re. Alkalin<: phosph:l.t::tse-conju­gated gOUt anti-moLl.se immLtnoglobLtlin IgM F(ab):. (Tago. Inc .. Burlingam<:, USA), in a dilution of 1:10.000 in PBS­Tween-20. \V.:lS used .::IS second step antibody (incubation..\.5 minut<:s at room temperatLtfel. B<!twe<!n each incubation step the blots were rinsed fivo:: times with 50 ml PBS-Tween-20 (O.l%). Phosphatases were visualiz<:d using a protocol from BI.:lke and coworkers (1988). For control immLtnoblots we Ltsed the unconditioned medium which was used to CLtlture the HNK-l hybridom:l cells. All otho;:r steps were identical.

Results

HNK-j immunoreactivity in the developing gur from E.f. till E12 We determined the HNK-l immunoreactivitv in the normal developing hindgut during (E4-E6) o:nd after neural crest ceU migration and colonization (E7-E14).

At E4. therl": ure few differentiated cell types present in the gut. The epithelium is multilayered and there are no layers of smooth muscle cells present. The enteric mesenchvme is surrounded bv a thin laver of serosal cells. Proximal to the cecal bulges, HN'K-l immuno­reactivity is present on the OLtter surface of serosal cells (Fig. 1). Underlying the serosal cells. HNK-1-positive cells are present within the mesenchyme. HNK-l immunoreactivity is also present underneath the serosal cells in th,e most distal segment of the gut. but a 1 mm segment distal to the cecal bulges does not contain any HNK-l-positive cells.

Fi~. 1. LongitUdinal paraffin section of tho;: postumbilic::tl E4 gut. Proximal to the cccal bulges (B). HNK-l immLtnoreactivity is present on the outer cell membrane of ~eros.:ll cells (single arrow) :lnd in the mesenchyme underneath (double arrows). 'In the c:lud.:ll gut. HNK-t immunoreactivitv is confined to cells in the se:oS:::t (arrowheads). The 1 mm gut segment distal to the cecal bulg';s does not contain HNK-l immunoreactivity (HNK-l-). x25.

111

T .. VI. Luider and others

,. '

., ,

EP' '.

':'~ .. ':~;: . ... -~~ ..

EP

',;.

:.',

Fig. Z. HNK·l·stained cryostat sections of E5 (A), E6 (E), E7 (C) and E14 (D) hindgut segmentS. HNK·l visualises the developing enteric ganglia. At E5 and E6. faint HNK·l immunoreactivity is present between aggregates of neural crest cells (arrows). At E7. neural crest cell colonization of the (hind)gut is compkted. At E14. HNK·L immunoreactivity in the myenteric ganglia is stronger than in the submucous ganglia. Compared with the submucous plexus. the size ot' the myenteric ganglia has increased considerably. ep. epithdium: m, myenteric plexus: nco.. neural crest aggregate: s. submucous plexus, Magniftcation 25x.

112

At E5. HNK-l immunoperoxidase staining of cryo­stat sections revealed some clustered positive cells in the mesenchyme of the periumbilical gut (Fig. 2A). In addition to the heo.yilv stained cell membranes of these clustered cells, we observed 0. wenker and disperse staining in the mesenchYme.

At E-6. clusters of HNK-l-positiye cells are located at the sites of the myenteric o.nd submucous ganglia (Fig. 2B). A:s in the E5 gut. we observed a disperse and fainter Hl"'K-l staining between the developing submu­cous ganglia.

At E7. the clustering of positive cells is more pronounced and one can distinguish between the mventeric and submucous ganglia (Fig. 2e). The m~senchyme was HNK-l negative. -

At E14. H:\,K-l immunope-roxidase staining revealed the relatively large myentcric plexus. located between the thin longitudinal and the circubr smooth muscle layer. and the relatively small submucous pleXUS. at the luminal side of the circular smooth muscle laver (Fig. 2D). The HNK-l immunoreactivity was located at the cell membrane of enteric neurons and their processes (immunofluorc$cence data not shown). The smooth muscle cells and the mesenchYme in the submucosa were not stained bY the HNK-'l antibody. Sometimes the apical site of the epithelium was HNK-l positive.

Four low relative molecular mass HNK-I-carrying plasma membranl! g!ycoproreins are present in £-1- bw not in £7 and EN hindgut To determine the molecular nature of the HNK-l antigens in the hindgut during the formation of the ente-ric nervous system. we produced two-dimensional gels for immunoblot analysis. In the plasma membrane fraction of E4 hindgut. we observed spots with relative molecular masses of 200. 130.44.42.27 (doublet) and 20 (doublet) x 10"' (Fig. 3A). The spot at 42 x to"; had a lower int~m:ity than the -4 x 10-'_spot. The HNK-l glycoproteins of 200 and 130 x 10-' represent various glycoproteins with different isoelectric points (pI range from 7.1 to 5.5). The HNK-l glycoproteins,of relative molecular maSSeS of -14. 42. 2i and 20 x 10 .... are single spOts. The isoelectric points of these proteins are 5~5, 5.5, and for both doublets 5.5 and 4.9. respectively. We did not observe any spot in control immunoblots.

In immunoblots ot" the plasma membro.ne fraction of E7 hindgut. we onlv observed intense SpOts in the high relative "inolecular mass range (200. 110-130 x tO J

) (Fig. 3B). The isoelectric points of these high relative molecular mass glycoproteins ranged from 7.1 to 6.5 for the 200 x 103 protein. and from 7.5 to 5.5 for the 110-

Fig. 3. Two-dimensional Ht""'JK-l immunoblots of plasma membranes isolated from different embryonal stages of the gUt (A. E4: B. E7: C, E14). Note th.: presenCe of low Mr HNK-l antigens (G--I4: G-4'2: G-27: G-20) in E4 hindgut and the absence of these mokcub in E7 and E14 hindrwt. HNK-l antigens in the high relative molecular mass redon ('200 and 110-130 x 1O~) are present boch in E .. and ~ E7/E14 hindgut. M. high molecular weight markers: S. plD.sma membrane staning material. The isodectric focusing range W;l$ determin.:d.by Carbamolytc markl::rs.

HN K-J anrigms and enreric nel.£fal crest

130 x 10"; glycoproteins, A characteristic "orion-like'" pattern of low intensity HNK-l glycoproteins w?~ found in the low molecul::tr range LH,/pI: :-\, 50 x 10.)/5.0: B, -1.0 x 1003 .'5...+: C. 30 x 1003 /5.8: D. 29 x 103./6. 1 )

facilitating the identification' ot" other low relo.tl\·e molecular mass glycoproteins. We did not observe ti;.,::

M

=-97--"

"'-... "-'II ,.-.

S M

,,-

,,-

S M

7:-__ '1 __ -:,

~:*"\0~.' 0-1,j' <t/'.

t 7.' ____ '1

t

'{ice

¥C

A

'.9

A

t ---- •. ,

113

T . .\1. Laider and ochers

114

HNK-J antigens and enteric neural crest

-. '.

"'.,- .. ~ .. . .-# ... , '" -~ ...... "

~:~ . ..... .- . .J< >

•

Fig. -l. (A) Confocal laser scanning image of cultured E';' hindgut. A byer of HNK-t positive mes<:nchymai cells inside circubr smooth musck layer is present (HNK-l mode 1). x16, (B) Detail of A. ><40. Ie) Confocal laser scanning image of a coculture of E..J. hindgut and the neural primordium. HNK-l-positive neur.:ll crest cdb or cnteric nCLlrons art: present in enteric ganglia (H:'\K-t mode ~). Note the :lbsencc of the layer of HNK-l-positivc mesenchym.:l! cells. x16. (0) Detail of C Th<: image was made at a different !evel tOrom that in C. ><25. (E) Cryostat section of cultured E-I- hindgut stained with the neuron specific :mtibody (E/CS), Note the abst!oce of immunoreactivity. (F) Cryostat secdon of cultur~d E-l. hindgut and vagal neural primordium staines! with EiCS. :'\Iote the strong immunoreactivity in the submucous plexus. x2S. E, epithelium:~. myenteric plexus; $, submucous plexus.

44.42,27 and 20 x 103 HNK-l glycoproteins present in E4 hindgut.

In two~dimen.sional HNK-l immunoblots of the plasma membra.ne fraction of E14 hindgut we observed intense SPOtS in the high ~dative molecular mass range (200, and 1l0~130 x ,lO") (Fig. 3C). The isoelectric points of the 200 x 10 .... P!otein ranged from 7.1 to 6.5 and for the 110-130 x H)" proteins the pI ranged from 7.1 to 6. The characteristic orion·like pattern of single HNK·l glycoproteins (A-D) was identical to that in E7 gut. The HNK-1 immunoblot findings are summarized in Table 1. From these results. we-conclude that E4 hindgut contains low relative molecular mass HNK·l­positive glycoproteins (0-44. G-42. 0-27. G-20) which are not present in E7 and EI4 hindgut.

HNK-] immunorcacrivity in chorioallantoic culwres of embryonic gut To test whether the disappearance of low relative molecular mass HNK-l antigens correlates with neural crest cen colonization of the gut. we cultured E4 hindgut without and with the vagal neural primordium

of stage HH 10 embryos. After the culture. we characterized the HNK-l~positive cell types and the HNK-l-carrving antio-ens.

Confocal "las;r sc~ning microscopy revealed that HNK-l-positive cells are present in cultured E4 hind­gut. These cells are located in (a) a circular layer of cells in the submucosa at the luminal side of the circular smooth muscle layer. and (b) in spots between the longitudinal and circular smooth muscle layer (Fig. 4A and B). The HNK-l-immunoreactive cells in the submucosa were connected to the HNK-l-positive spots at the site of the myenteric plexus by HNK-l-positive fibers. We will refer to this pattern of HNK·l immunoreactivity in cultured E4 hindgut as HNK-l mode 1 immunoreactivity. It is important to stress that the fiNK· I-positive cells in the submucosa develop in cultures of gut segments that do not contain neural crest cells or any other HNK-l-positive cell types at the time of explantation.

Coculture of E4 hindgut and the neural primordium resulted in an apparently normal enteric nervous system in the transplant (see also LeDouarin and Teillet. 1973;

115

T. ,'vi. Luider and ochers

Table 1. HNK-i anrigens (':.v1r x]rY) in plasma membranes of hindgw

Aganglionic hindgut G::mgliomc hindgut

~ormal Cllitured C(X:ultured ~ormal Cultured C(X:uilured :-<ormul E4 E4 E"-:-;P E7 E7 E7"':-<C W

:00 '00 :00 '00 :00 :00 :00 130 13<) 110-130 110-130 110-130 130 130 -" .w 4" 40 07 :0

mode 1 mode ::: mode ~ mode: modo: : mode :::

E. day of development: :-<P. neural primordium: yrode land:. type of HNK-t immunor.:octi"ily I~<:'O;: text).

Allan and Newgreen. 1980). The neurons and the neurites in the enteric ganglia were rINK-l positive (Fig. 4C and D). We will refer to this immunohistologi­cal pattern of HNK·l immunoreactivity as HNK-l mode 2. Culture of E7 hindgut also resulted in HNK-1 mode 2 staining.

In quail-chick chimeras. sacral neural crest cells migrate along the dorsal surface of the gut and give rise to the ganglion of Remak during E4 through E6 (LeDouarin and Teil.tet. 1973). To exclude that the HNK-l mode 1 staining in cultured E4 hindgut is due to neural crest cells that have migrated from Remak's ganglion through the serosa. we dissected the E5 hindgut and removed Remak's ganglion (it was not possible to remove Remak's ganglion from E4 hind­gut). After one week of culture. HI"'K-l visualized the layer oE cells in the submucosa. HNK-l mode 1 (data nOt shown). Thus HNK-l mode 1 is not due to sacral neural crest cells that have migrated from Remak's ganglion.

To investigate whether the HNK-l mode 1 cells have nl!uronal characteristics_ we pt::rfonnt::d immunohisto­chemistry on cryostat sections with the E/C8 antibody (Fig. 4E). We did not find E/C8-positivc: cells or cells with a neuronal phenotype in the cultured E4 hindgut. In addition. there were no neurofilament-positive cells in cultures of E4 hindgut (data not shown). In contrast. HNK-l mode 2 coincided with E/CS-positive cells in the enteric ganglia (Fig. 4F). HNK-l mode 1 did not coincide with immunore:lctivity with antibodies specific for three characterized HNK-l antigens (N-CAM.

Day of dissection

E4

Successive bowel segments

IIU

chicken integrin_ tenascin: data not shown). We determined ';hich segments of the gut exhibit HNK-l mode 1 by culturing -successive segments of postcec:ll bowel oE E5. E6 and E7 gut. As is shown in Fig. 5, HNK·l mode 1 reactivity is present in culturcs of the most distal gut of E4 through E6. By E7 all cultures of the gut show HNK-l mode 2 immunoreactivitv. The cultures of the distal E4 through E6 gut did not contain enteric ganglia. thereby representing agunglionic gut. Therefore. HNK·l mode 1 immunoreactivitv is con­fined to aQ;angiionic segments of the gut, while HNK-l mode 2 immunoreacti~ity is related -to ganglionic gut segments.

Two low relative molecular mass Hj\'K-l carrying plasma-membrane glycoprou:ins are present in cultures of £-J gw bw nor in (co)culwres of £7 gw and neural primordium We cultured explants of E4. E7 hindgut and cocultured E4 hindgut and the neural primordium until both types of expbnts had reached the age of 11 devdopmental days.

In immunoblots of the plusma m'embrane fraction of cultured E4 hindgut we observed HNK-l-positivc spots with relative molecubr masses of 200.130. -l4 and 42 x 103 (Fig. 6A). Thc rangl! of isoelectric points of these glycoprotcins is similar to that observed for HKK-l antigens in explanted E4 hindgut. The orion-like distribution of HNK-l antigens A through D facilitated the proper identification of HNK-l antigens in the

R

IIU HNK-1 MODE 1

"'-c E5 HNK-1 MODE: 2 HNK-l MODE 1

llU /

E6 HNK-l MODE2 HNK-l MDOE2 HNK·l MODE 1

IIU C E7 HNK-l MOOE2 HNK·' MDDE2 HNK-l MDDE2 HNK-1 MODE2

116

Fj~. 5. HNK-l renctivi(v in cultures of succl::ssivl:: s~gmen(s of postcec::tl bowel of Eg. E6 ::tnd E7 gut. R. rl::ctum: C. cecum: U. umbilicus.

• A 200- :1 t

A

<3-

25- ' ..

,,- .... . - • t t

7,''----- or ------' ....

200- :1 07-

B

.. - .. 0'

·A

..,- II o

,.- .. 1a-~

t 7,'

or t

200- __ E 'I gr_. r

68- .. A

43-

~-, ,e- .,.'

t 7,1

t ~------DI------~4~

t

HNK·l amigens and emeric neural crest

c

A

----~ t or --------0,$ ... ~-

D

t 01------'

'$

Fig. 6. Two-dimensional HNK·l immunoblots of plasma membranes obtained from (c:o)cu.lmfed E-l. hindgut (A). E..). hindgut and neural primordium (B). E7 hindgut and neural primordium (C). neural primordium (0) and E7 hindgut (El. The two proteins G·"'2 and G-.w. disappear during neural crest cell co!oni7~tion (eoeulmre of E-l- hindgut and neural primordium). Accurate lOc:.l.lization of G·"'2 ~:lnd G·44 is possible by the enteric HNK·l SpOts A through D, M. molecular weight markers: S, plasma membrane starting material.

117

T. .W. Luider and orhers

lower molecular range. The doublets at 27 and 20 x 103

had disapp~ared. But the HNK-l glycoproteins of -14 and 42 x 10-' were still present.

In immunoblots of the plasma membrane fraction of cocultures of E4 hindgut with the neural primordium. we observed intense spots in th~ high relative molecular mass range (200, 130-110 x 10'» (Fig. 6B). The orion­like distribution of HNK-l antigens A through D was also present. The G-42 and G--14 H~K-l glycoprmeins were absent. Thus during the coculture either the expression of the HNK-l epitope on G-42 and G-M or the expression of these HNK-l-bearing proteins disap­peared. Neural crest cell colonization did not yield additional HNK-l-positive spots.

To test which HNK-l glycoproteins are expressed by the neural primordium. we cultured it in combination with E7 hindgut or alone (Fig. 6C and DJ. In immunoblots of the plasma membrane fraction of cultures of neural primordium. we observed HNK-l­positive bands in the high relative molecular mass range (200. 130-110 xlO) (Fig. 6D). The orion-like distri­bution of H:-\"K-l antigens A through D that is normally seen in plasma membrane fractions of the gut was not observed. Furthennore. the G-42 and G--l4 were not detected.

In immunoblots of the plasma membrane fraction of cultured Ei hindgut, we observed a string of Jntense spots in the high Mr range (ZOO. 110-130 x 10-') (Fig. 6E). In the low ,1./r range. we observed HNK-l antigens A through D. The HNK-1-positive spots G-42 and G-+1-seen in (cultured) explants of E4 hindgut were absent. In cocultures of E7 hindgut and the neural primordium, we found a similar picture to that found in cultures of E7 hindgut alone (Fig. 6C and E). The HNK-l antigens in plasma membranes of (co)cultured E4 and E7 hindgut are summarized in Table 1.

From these experiments. we conclude that G-..J.Z and G-44 are present in (cultured) explants of E4 hindgut and absent in (cultured) explants of E7 hindgut. The spatiotemporal expression of G-42 and G-M coincides with neural crest cell colonization in the hindgut.

Discussion

HNK-l-positive enteric mesenchyme and rhe formarion of enreric ganglia in rhe poswmbilical gut HNK-l immunostaining of E4 hindgut sho>,\'S that HNK-l-positive cells ar"e located anterior to the cecal bulges and in the most distal colo rectum. The anterior fP.tK-l-positive cells could reflect the vanguard of migrating vagal neural crest cells. However, the presence-of neural crest cells just proximal to the cecal bulges does not correspond with the time of arrival of neural crest cells that has been reported in earlier studies (LeDouarin and Teillet, 1973; Allan and Newgreen, 1980). According to these authors. vagal neural creSt cells migrate in the preumbilical gut at E4 (stage 24). They reach the umbilical region by ES. and the cecal region by E6. The presence of H~K-l-positive cells in the most distal segment of the hindgut at E4 is in

118

agreement with the studies of Pomeranz and Gershon (1990, 1991),

We found that HNK-1-positive cells develop in cultures of HNK-l negative explants of E4 hindgut. These H?"K-1-positive cells are distributed as a mes­enchymal cell layer at the luminal side of the circular smooth muscle layer (HNK-l mode 1). Thus, even if HNK-1-positive vagal and sacral neural crest cells are not present within a hindgut-segment. H?-:K·l-positive cells still develop. This suggests that the WK-l­positive cells in HNK-1 mode 1 do not derive from the neural crest. The existence of H0'K-l-negative neural crest cells in stage 24 gut is an objection to this assumption. However. this is unlikely since it is generally accepted that the majority of sacral neural crest cells are Hi:"K-l positive (Pomeranz and Gershon. 1990). In view of the distribution of the HNK-1-positive mesenchvmal cells in the submucosa and in the myenteric region. we surmise a splanchnopleural mesoderm origin. An endodermal origin for these cells is not likely but cannot be excluded. -

The HNK-1-positive mesenchyme in cultured E4 hindgut seems to be organized in a network. The submucosal layer of HNK-1-positive cells is connected to the HNK-l-positive cells in the myenteric region by HNK-l-positive tracts. Epstein and coworkers (1991) observed a neural crest-derived HNK-1 positve network in the foregut which initiates the formation of the enteric nervous system. We found a ffi'\K-l-positive network in the hindgut which does nor derive from the neural crest. It could well be that the fonnation of the enteric nervous system in the foregut and hindgut may be mediated by different mechanisms.

,Vew members of rhe L2/HNK-I family Although the HNK-l epitope is present on a family of cell adhesion molecules. it is surprising that a wide array ot HNK-l antigens is present in the developing gut at particular developmental stages. Most of the cloned members of the L2/HNK-l familv of cell adhesion molecules have relative molecular -masses over 100 x 103

• The high Mr HNK·1 antigens in the plasma membranes of the developing gut could be known members of the HNK-l familv of adhesion molecules such as the neural cell adhesion molecule N-CAM (M r =200. 180. 160. 140 x 10"3), the f3 subunit of the fibrogectin receptor and the laminin receptor (Mr =13S x 10").

HNK-l af1tigens with low relative molecular masses «100 x 10'» are more abundant at early embryonic stages compared to later stages (Canning and Stem. 1988). We detected low relative molecular mass HNK-1 antigens in immunoblots of two-dimensional gels of plasma membranes ot early (E4) embryonic gut (G--I4. G-42. G-Z7 and 0-20). The HNK-1 family of adhesion molecules has grown substantially and several of the cDNAs encoding tor the protein backbones have been cloned and sequenced. Only two members of the HNK-1 family with relative molecular masses lower than 100 x 1O~ have been iqemified: the myelin protein Po (Mr=19.6-26.5 x 10 .... ) (Lemke et aI., 1988) and an

acetvkholinesterase of Electrophorus electric organs (.'.1r -o.pproximatdy 70 x 10» (Bon et al" 1987). Thus the relative molecular masses of the H:'\K-llntigens G-42 and G-M do not resemble known H::--:K-l a;tigens. There is little knowledge about the proteins that co.rry the HNK-l epitope during early vertebrate develop­ment. However. most. if nOt all. members of the known members of the L2/H:-\K-1 family play :l role in cell adhesion. We surmize that G-":'2 and G--+4 reo resent two unidentified H:;..iK-l-co.rrving cell adhesio'n mol-ecules. - -

G-·/2 and G-.J.+ in E-I hindgw disappear dwing neural crest cet! coloni:::.alion ::-:eural crest cell colonization of the hind'.!;ut occurs during E4 through E7. We dt:;tected low .VIr HNK-l antigens in the gut prior to neural crest cell coloniz­ation. T\vo of these. G-20 and G-27, are present in explants of E-.+ hindgut and disappear during culture. G-42 and G-44 proteins :lre present prior to neural crest cell colonization but they disappear after neural crest cell colonization. both in vivo and in cocultures,

An important issue to resolve is which HKK-1 antigens cause HNK- L mode 1. Due to the lack of additional markers for G--'+2 :lnd G--14 we cannot ascribe HNK-1 mode 1 exclusively to these two proteins.

In sections of cultured E-'+ hindgut most of the HNK-l immunoreactivitv is located in H3.:K-l mode 1. whereas in immunoblots of similar cultures the most prominent HNK-l-positive SpOts are found in the relative molecu­lar mass range of 200 and llO-l30 x 10". Because high relative molecular mass HNK·1 antigens are present in (cultured) E7 and E14 gut and in cultures of the neural primordium. it is not unlikely that they represent neuronal antigens. The presence of these neuronal HNK-1 antigens in E4 hindgut can be ascribed to the presence of extrinsic nerve fj-bres or Remak's ganglion, Another possibility 'I~ that the HNK-l-positive mes­enchymal cells contain these neuronal HNK-1 antigens and that the expression continues after neural crest cell colonization.

HI"'K·l-positive cells and antigens are essential for gastrulation and the development of the mesencephalic neural crest in chick!;:n t::mbryos (Stern and C:lnning, 1990: Bronner-Fraser. 1987). These findings, takcn together with the adhesive characteristics of HNK-l antigens. suggest that the Hl':K-l antigens in E-'+ hindgut might playa role in adhesive or in repulsive interactions with enteric neural crest cells. As such, the HNK-l-positive mesenchymal cells might playa role in the initiation of the patterning of the enteric nervous system in the hindgut.

The authors thank Mrs IIse van Haperen-Heuts and Mrs Sandra van Galen for technical assistance. Wc th.:ln.k Tom de Vries-Lentsch for a:;sist.:lnce with photogwphy and Ko Hagoon for ediwri.:ll help. This study W.:lS supponed by the Sophia Foun.dation for Medical Research. Th.M.L. was supponed by a research gwnt of :vJ:edigon. II'\WO gr.:lnt no 900-522-073) .:lnd ,hI:: ~etherJ.:lnds Digestive DiSeases Foun-dation. (gront no WS 89-::1l. -

HNK-J antigens alld enteric neural creSl

References

-\bo, T. und B:llch. C, :"I.119$11. A differentiauon Jntl!!en ot human :-;K Jnd K cells idendned by J monoclonal nntlbody' IHNK-[l. J. /mm.wwl. IZ7. 10~"·1O~9.

Allan. L J. lind \'ewgreen, D. F. I 1980 J. Thc on~m Jnd dlfferentintlon of <:ntenc neuron~ of the intestine of the chick embryo. Arna, J AlIai. 157. [}7·1.5".

Blake, ~l. S" Johnston, r;:, R., Russd-Jon~"'!, G, J, and Gotschlkh, E. C. 119&l). A rnpid nnd 5en~itive method for detection of nlklline ph05phat~se. Anal. Blodlern. 1.36. lS7·li9

Bon. S .. ;"[~f1ah, K .. ;..ru.~sct, F .. Gras~i, J. and :"[as$ouli':, J, (1987). An Immunodobu!in ;.....( monodonal antlbodv. reco~nIZln~ a ,ubset of nCetylcholine~ternse molecules from' elect~c organs ot £/,'Clrophorm and Torpedo. belon~$ to the HNK·1 nntl­c~rbohvdrnte f~milv. J. Se!jracl:~rn, .. 9. 17:0"li:H

Bronner.Fr.l..~cr, :"1. '(1987). Perturbutlon of cranial neurnl crest ml\1.rnlion bv the H:\,K-1 nnClbodv. D~"I 8'01. l~. 3~1·31.

BUl1:~r, 0 .. S(mon, :"1., Perrulsscau', G, ;J;nd Steck, A. J. (19QO). The epitopel,) recognized by H;..iK-1 ~nClbody nnd IgM pnraprotCIn:n :1europathy IS pre.,ent on severnl ;-:·linked oligo,accharlde ~t:'UClUre,> on human p" ;:md myelin ns.wduted f'l:.coprotein. J. .veurochern. $4, IS6~-157S.

Cannin~, D. R. nnd Stern. C. D, (198$). Ch;lnge> in thc expre""lon or the cnrbohydrutc epitope H:-':K·! l,SOl:lated with me~odcrm lnducdon in the chick embryo. De"dopmenf 10 ... 643-655.

Chou, D. K. H .. lIyas, A, A .. Ev~ns. J. E" Costello, C .. Quark'S, R, H. and Jung~lwnln, F, S, (1986). Structure of ,ulfaced glucuronYI glycoiiptds in the nervous system reacdng IVlth HNK-i Jntibody and ~ome IgM paraproteins in neuropa~hy. J. BioI. Chern. ~61 (25). 1li17·~

Chou, K, H.. Ilyn~, A, A •• Evuns, J, E" Quurlcs, R. H. and Junt::nlwala, F. S, 1191'S). Structure of a glycolipid rcacting With monoclonal IgM in neuropathy nnd with H:-':K·1. Hior:ilcm. Biophys. Res. Cornrnun. I~S, 383-3SS.

Clment, G .. R~ler. A., Lctoum<.'nu, P. C. and VVeston, J, .-\. 1(986). !dentirkatton of In interfiiamcnc '~lo."oocinted protem. :-:APA·73. binding to different filament types Qt different stJge5 dutlng neurogenesis. J. Cell Bioi. IO~, 246·251.

Clment, G, :lnd \V~'Ston, J."\, I 19K!). Early appenranco.: in neuml crest and crest·dcrlved cells of ~n Jnti~eni<: determinant present in JVtun neuron,. De,'1 BIOI. 9.3. 3S5-367.

Dennis. R. D .• Antonicd;, H .• Wiegandt, H. nnd S<:hachn"r, ;\1. I 1985). Detection of the L~!HNK·l carbohydrnte epitope on glycoprotein~ and aCidic g!ycolipicL~ of the inscct Calliplwra ,·Icma. J .. V,~uror:hern. 51. 1490·1496.

Epstein, )1. L.. Poulsen, K, T. and Thiboldcau'>, R, 119(1). Formatton of gangll~ in the gut ot the chick embryo. J. Camp .Veurol. 307, 189·199.

Faissner. A. (1987). :vronoclonnl nnttbodv detect, c;:rbohvdrate

.~~~,r~~~:.erz~~.n~~~3~,n 3~~~3~~urine cell 'Jdhe5ion molecule LL

Feizi, T, {l985). Demon'>tration bv monodonn! antlbodtes thnt carbohydrate ~tructure$ ot' glycopr~teins and glycoiipld~ Jre onco­developmentul antigens . . Vawre 314. 53-5;.

HJmburgt'r, V. nnd Hnmillon, H, L, 1 19S1). A ,eries ot norm~li stn~e~ In thc development ot the chick embryo, J .• '-torpilol. SS. +9-67.

Hofl'mon, S. nnd Edclmon. G. :"1. (1983\. Kinetics ot' homophilic binding by embrvonic nnd adult form~ of :he neur;:! cell adhe.,ion molecule: Proc. ;"/arl. Acad. Sci. USA SO. Sil!~-5766.

HotTman, S. ~nd Edelman, C. :"1, 1 1(137). r\ prot¢o~lycan With HNK·1 anw:enlC determtnnnts is a neuron-os~ocluted lit::nnd for cytowctin. Proc. ,':ml. Acad. Sci. CSA &l, ~S2J-~527.· .

Holley, J, A. nnd Yu. R, K. (19137). LOI:alizntton of ~lycoconjugntes recognized bv the HNK·I ant'lbodY 'In mou~e and chick embrvos dunng early neurnl development. De.' .. Vc"lfOSCI. 9, lOS·1l9 ..

Imnmufn, T. and :"litsul. Y. {19S7). Hepar~n '>I.dfate and hepnrin n.s a potentiator nnd ~ ~upprc>sor of growth ot' norm~d lnd trnn,formed yuscular endothelial cdb. E.rp. Cdl R~s. 111.92-100.

Keilhauer, G" Fnissn~r, .,\, nnd Schachner,;\1. 1(985). Differentinl inh:i:>ition ot' neuron-neuron, neuron-n,trocvte and astrocyte­Jstrocyte adhe,'on bv U. L:. nnd :-':"CAM nntlbodics. 'varur<' 316. 7;;$-;30. .

Kru~e, J .. '/;!a·llhammer. R .. Wern<."<:ke, H" FD.\ssn~r, A .. Sommt'r, t,

119

T. 1'vI.. Luider and Olhers

Goridb ... C. ;lnd S<:lU1dln~"!" .. "1.1. tI9...~1. :-f<:ur;tl .;dl ndhe,ion molecul<:5 Jnd myelin-n~"odUled Slycoprotein ,h;tre ..I common carbohydrate moiety reco~nlzed bv monodonal .)[\tjbodie~ L~ nnd HNK.l .. Vatllr( 311', lS.3.iss. .

KUn~mund. V •• Jun~alwala. F. B .. Fbher. C .• Chou. D. K. R •• Kdlhuu<!f. C. and Schachner. :>1. (19881. Thc L: H:'\K·I carbohvdrate 01 neural.;ell adhe,wn mok:cule5 l~ lrl\olvcd in cell lrlte[aciiol1~. 1. Cdl. Bioi. 106. :13-~:3.

Kuratani. S. C. {19"911. Alternnte eXp[e'~IOn of d:e H:--iK·I epltope 1[\

rhombomere, of the chick~n embryo. Dc,,1 Bioi. l.1J. :t5-:19. LcDouarin. :"\. ;"1. and T"ille!. :>1 • .-\. (1973). The mi1![atlon of neural

ere5t cells to the wall of the di~e~tive tract in :\vlan embrYo. J. Embryol. Exp .. Horpnol. 30 . .31:"S .

Lemke. G .. Lnmnr. E. nnd P:lttcf'>On. J. (191:'8). holation and ;;tn;tlysi~ of the gene encodin~ peripheral myelin protein zcro. -"("ron 1. 73-83.

:'-lu~-da. T .. Balakrishnan. K. and )Olehdl. S. Q. (1983). A ,imple a:1d rapid method for the prcparanon of pb,ma membr;tne~. Biocilim. Blophys. Acra 731. 115-1::'0.

)oleijers, J. H. C .• Tibboel, D .. Van der Kump • .-\. W. :>1.. Van H:lpercn·Hcuts. C. C. ;"1 .. KlUck. P. :lnd :>(Olen(Ulf. J_ C. (1<i87). The influence ot" th~ ~tJ.l:e of dift"ere:1t1al'On of the ,,\.It on the mi"r:ltion of neur:!l c~lb. An experimental nppronch of Hirscn5pruns',; dise~e. P(diurr. Res. ~l. 131,135.

:-"ktcalfc, W. K .. ;"Iy~rs, P. Z .. Trcvarrow. B .. Bn."'S. )01. B. and Kimmel. C. B. (1990). Prim:!ry neurons thD.t expre~, the L::'jH::-:K-l carbohydrate dudng early dev~lopment in the zcbrutbh. Devdopm(nI 111} ... 91-504.

;"Ilkol. D. D .. Wrabetz. L .. )o!:1rton. L. S. :lnd StcfanS50n. K. (1988). Devclopmentnl changc~ in the molecular weight, ot polypept;de~ in the human CNS that carry the HNK·l epltope :lnd bind Pha:;(olli$ ,'"I::(lfIS lectins. 1. .'.fcrlfocil(m. SO. 10::. .. ·19:$.

l"ewj!reen. D. F .. Powel. M. E. und :>Ioser. B. (l()lJO). Spntiotcmporal changes in HNK·!jL1 glycoconjugate~ on avian ..-:mbryo ,.,omite and neural cre$t c~l1s. D.-vi Bioi. 139 (1).100-120.

:"iordJand<'r. R. H. (1989). HNK·l m.,[ks enrli<;~t ~xom'! out~ro\\ith In Xenop!<:;. Dey. Brow Res. SO. 1 .. 7-153. ~

O'Shann~'!>"!."y, D. Jo, Willison. H. J .. Inuzuka. To, Dohcr~cn.,"\1. J. und Quarles. R. H. (1985). The ~peei<::~ dbtributioll ot" nervou~ ~y,tem ;tnti<;cns thut re;tet with anti·myelin-uS5oc;atcd glycoprotein antibodie~. J .. Vt'liroimmmlOl. 9. 255-::'68.

Pl'Shcva. P .• Horowitz. ,\. F.llnd Sehnehncr.:-"1. (1987). Intcgnn. the cell wrface receptor for ~broneetin and laminln. expre~~e~ the L2/HNK-l ~nd L3 curbohydratc ~tructt.lre~ "bared by ~dhe~ion moit:cule,;. ,V.-UfOSCI. Lerr . .$3. 303~306.

Poltornk. M .. Fre<.>d. W. J. and $chnch·ner. ;"1. {19891. Expression of ~dl adhesion mole~ule~ from the L::'iHNK·l bmily in ~er~bcllar iso\,:r:ltlS in mice. Brain Research 4SS. 265·27 ...

Pom ... 'ranz. H. D. and Ger-hon, ;"1. D. 11990). Colonization of the

120

J.VI:ln hind)!UI by celb (knved from the ~nernl neural ~rest. D.:vl Bioi. 137. }78-}':14.

Pomerunz. H. 0 .. Rothmun. T. P. and Gershon, ;"1. D. (1991). Coloniz:ltion of thc post-umbilical bowd by cells de:wed from the 'acrnl :1eural cre~t: direct trueing ot cell migrudon u5ing un intercalJting probe .. md a r<,plication·de~cknt retrovlru~. D~\"elopmt'!11 Ill. (,.J7·655.

S:ldoul. R •• Hirn, )01.. IkagO'ilini-Bnzin_ H .. Rou)::on. G. nnd Goridls. C. {198.3). Adult and embrvon.ic mou~e neur:ll cell ~dhe~ion molecules have different bi~di[\g properties. SarUff! 30-1. 347·3-+9.

S<:hwartin)::, G. A .. Jun)::!lI .... :Jla. F. B .. Chou. D. K. H .. Boyer, A, ;"1-nnd Yamamoto. :>1. (1987). Sulfated "lucuronic acid cont::l.inin~

glycoconjusot<:~ Jrc tempor~lly and ~pn!ially re!>ulated Jnw;ens m the developl~S m:lmm:lli:ln nervou, ~y~tem. D( .. I BiOI. \:::0.65-76.

Serb~,.jzija. C.:"i .. Bur~:.ln, S .. Fra.~er. S. E. and Bronner-Fraser. )01. 119(1). Vit;).1 dye I:lbcl;n~ demOn~trules J ~aeral neural erest ~ontrlbutlon to the en!ai~ n"f\'ou, ~ystem of ~hick and mous<: embryo,. Dc,.e!opmc·nr 111. 857-867.

Shashoull. V. E .. D:lnld. P. F .. )oloore,:-"1. E. and Jun)::ulwalu, F. B. 1 1(86). D~mo~~tratlon of glucuronic a~id on brain glycoproteins which reuct With H:\K-l ;tntibody. Bioch~m. Biopnys. Rt's. Commull. 13$.90::'·909.

Sorkin. B. C .• Hoffman. S •• Eddnmn. G.:-"1. and Cunnin)::hnm. B. A. (19S-l-1. Sulfntution Jnd pho~phorylution of the neurnl cell adhe~ion molecule. :--i·CA:Vr. Sci':llc,: :::::S. 1 .. i6-I..:78.

St~rn. C. D. und Cunnin)::. D. R. (1990). Ori1!m of ccn~ ,;:ivin" rt5e to

me~oderm Jild endOderm In chick embryo . . Varur~ J.i3. ::73-275. Thorpe. S. J .• BeH:lirs. R. and Fcizi. T. (t9SS). Developmental

pntlernlng of c;;Irbohydro.te Jntlgcns during ~arly embryogenesi~ of the chick: expre~,ion of ;;Intigen~ ot the poly-;-.r-~cetyll:!cto~mIrlC ,erie,. Dn-r.lopmenr 10:::. 193·::'10.

Tuckl'r. G. C . .-\oyamll. H .. Lipinski. :-,.[ .. TUfs;r. T.!ll1d Thiery. J. P. r 198J). Identicul rcnctl'"itv of monOclonnl ontibodies HNK·l ~nd :\C-l: conservation Irl ~eri:ebr;tte5 on cell!> derived from the neural primordium and on ~ome leukocyte,. Cdl Dlff~renriailon 1 ... ::'2.3-DO.

Tucker. G. C .. Oment. G. and Thicry, J. P. (l986). Pothwoys of nvinn nc,lrol crest cell migration in the dcveloping gut. D(vi BIQI. ll(j. .. 39-.. 50.

Tuck<'f, G. C .. Dclatue. ;\1.. Zada. $ .• Boueau!. J. C. and Thiery. J. P. ([988). Expre~slon of the H>J"K-1.':-IC-t epilOpe in ently vertebrntc :1~urogenesl'. Cell Tissue Res. ~51. -1.S7-b5.

Vinc~nt • .\II., Dubnnd, J~L, ~nd Thiery. J. P. (19831. A cell5uriac~ dctenninanr expre~~"d early on mlgrati:1': avian neur:ll cre,t cells. Dev. Brain Res. 9. :!3S-138.

(ACcepted 18 :'.Jarch 1992)

CHAPTER 3.7.

Pattern of malformations and dysmorphisms associated with Hirschsprung disease: an

evaluation of 214 patients

Marianne W. Van Dommelen. Mrujo J.H. Peters-van der Sanden, Jan C. Molenaar, and Carel

Meijers

Abstract

Several lines of evidence suggest the involvement of genetic factors in the pathogenesis of

Hirschsprung disease. As an initial step to identify these factors we detennined the incidence

and nature of associated anomalies both in long-segment and short-segment Hirschsprung

disease. We retrospectively examined the charts of 214 patients with Hirschsprung disease

admitted to the Sophia Children's Hospital between 1970 and 1992. We -found that short­

segment Hirschsprung disease was present in 170 cases Cd': ~,4: 1), long-segment in 44 (d'

: ~. 3: I). The overall incidence of associated anomalies was 27.6%. We found associated

anomalies in 39 patients with short-segment (22.9%) and in 20 patients with long-segment

disease (45.5%). We distinguished five classes of Hirschsprung disease depending on the

presence and nature of the associated anomalies. The first class entailed isolated cases of

Hirschsprung disease. The d': Ej1: ratio was 4.7 : 1. with no difference between short- and long­

segment disease. The second class entailed cases with Down syndrome (7.6%); these patients

had short-segment disease and were predominantly males. The third class entailed' syndromic'

cases, the incidence of which was highest in long-segment disease (20.5% versus 5.9% in

short-segment disease). The d' : ~ ratio in 'syndromic' cases was 1 : 1. In the fourth class

Hirschsprung disease was associated with craniofacial dysmorphisms. and in the fifth class

with one other anatomic malformation. We conclude that short-segment and long-segment

Hirschsprung disease do not differ markedly with respect to the pattern of associated

anomalies. sex ratio. and familial occurrence. It might well be that isolated forms of short­

and long segment disease are variant forms of the same genetic defect. The association with

Down syndrome seems to be confined to short segment Hirschsprung disease.

121

Introduction

Hirschsprung disease (HSCR) is characterized by the absence of enteric neurons and the

presence of hypertrophic nerve trunks in the distal digestive tract. Over the last 30 years.

many case reports and several large cohort studies have established clinical, pathological and

genetic heterogeneity for HSCR. A positive family history has been described in

approximately 7% of all cases [Kleinhaus et al .• 1979]. Genetic study of HSCR has proven

difficult. because of the limited availability of large pedigrees. Until fairly recently. patients

with HSCR hardly gave rise to offspring. The high number of sporadic cases and the fact that

the disease is four times more frequent in boys than in girls suggested a sex-modified

multifactorial mode of inheritance, involving multiple genes. Badner et al. [1990]. however. showed that. while for short segment (SS)-HSCR the inheritance pattern was equally likely

to be either multifactorial or due to a recessive gene with low penetrance, for long segment

(LS)-HSCR the mode of inheritance was most compatible with an autosomal dominant gene

with incomplete penetrance. HSCR may be associated with other malformations, but the

reported incidence ranges from 5.3% to 29.8 [Ehrenpreis. 1970. Spouge and Baird. 1985.

Ikeda and Goto. 1986]. Many of these studies were limited by the small size of the

populations studied. and attention was mainly focused on major malformations that required

surgical repair, not on detailed dysmorphologic description of HSCR patients. Several of the

combinations of HSCR and associated anomalies acquired a McKusick index number,

representing Mendelian Inheritance in Man (for O:MIM: numbers see Table I).

Well-defined patient populations are a prerequisite to study congenital malformations

Tahle I:

1

2

3

4

5

6

7

8

9 10

11

122

McKusick Index numhers of Congenital Malformations of the Enteric

Nervous System

Hirschsprung disease (HSCR)

HSCR with ulnar polydactyly, polysyndactyly

HSCR with type D brachydactyly

HSCR and Ondino's curse

HSCR and microcephaly and iris coloboma

HSCR with pigmentary anomaly

249200

of the big toes.VSD 235750

306980

HSCR and hypoplastic nails and dysmorpbic features

HSCR and Waardenburg syndrome type 1

209880

235730

277580

235760

193500

235740

202550

243180

HSCR and polydactyly and renal agenesis

Total intestinal aganglionosis

Intestinal pseudo-obstruction

using molecular genetic techniques. As an initial step to study possible pathogenetic

mechanisms of one or more varieties of intestinal aganglionosis, we determined the sex-ratio

and the incidence of associated malformations and dysmorphisms in a large population of

HSCR patients, and characterized these. We wondered whether we could distinguish SS- and

LS-HSCR based on these parameters. Establishing the presence of associated malformations

might help in syndrome delineation. Even clinically insignificant anomalies may bear major

information value for diagnostic and epidemiologic purposes.

Patients and methods

We collected the files of all 227 patients with histologically proven HSCR admitted to the Sophia Children's

Hospit:l.l. between 1970-1992. The data abstracted from the files included: (1) sex: (2) length of aganglionic

segment (pathology reports): (3) presence of associated malformations and/or dysmorphisms: (4) presence of a

family history of aganglionosis or other congenit:l.l. malformations; (5) causes of death in the 17 patients who

died. Not all parameters were available in each file. In 13 patients. we were not able to trace the length of the

aganglionic segment. These patients were excluded from further analysis. A questionnaire was sent to patients

admitted between 1970-1985 (patients indicated with A in tables). asking for familial HSCR and for associated

anomalies in the patient. From 1986 till 1992 (patients indicated with B in tables). 33 of 59 patients had been

evaluated by a dysmorphologist.

Definitions: First we classified the patients according to the length of the aganglionic segment. measured from

the internal anal sphincter. We will refer to HSCR involving anus. rectum and part of the sigmoid colon as SS~

HSCR. Long~segment HSCR entails the colon proximal to the sigmoid colon. whereas total colonic aganglionosis

(Zuelzer Wilson disease) involves the entire colon. Sometimes the terminal ileum is also involved. but the

aganglionosis rarely involves the entire gut caudal to the duodenum (total intestinal aganglionosis). We will refer

to long~segment HSCR. total colonic aganglionosis. and tot:l.l. intestinal aganglionosis collectively as LS-HSCR.

We defined major and minor malformations as anatomic abnormalities that are present at birth (e.g.

pulmonary artery stenosis. renal agenesis. or Meckel's diverticulum). We also included disorders such as

psychomotor retardation. deafness. and central hypoventilation as malformations in our analysis. Dysmorphisms

are defined as developmental deviations from the usual morphological form. with an antenatal origin. The term

anomalies will be used to indicate both malformations and dysmorphisms. Patients with only one minor

malformation or dysmorphism. such as a sacral dimple or abnormal handcreases. were scored as having no

associated anomalies.

Statistical analysis was performed using the X2 test. using SPSS software.

Results

SS-HSCR was most common (n = 170, 79.4%). while LS-HSCR was present in 44 cases

(20.6%) (including 11 cases with total colonic aganglionosis and 3 cases with total intestinal

aganglionosis). We did not ascertain cases with ultrashort aganglionosis or skip lesions. Of

the 170 SS-HSCR patients, 136 were maIe and 34 were female (d' : ~, 4: I). The d' to ~ ratio

123

for LS-HSCR was 3 to 1 (33 males to 11 females). The distribution of the different lengths

of the aganglionic segment and the sex ratio are depicted in Figure 1. These two parameters

are in agreement with frequencies reported in other series [Kleinhaus et al., 1979, Ikeda and

Goto, 1984]. This indicates that Ollr patient population is a representative group of cases with intestinal aganglionosis.

The overall incidence of associated anomalies was 27.6%. In our A series (admitted between 1970-1985), the incidence of associated anomalies was 19.8%, and in our B series

(admitted between 1985-1992) 48.3%. This difference in incidence is statistically significant

(p < 0.001)

Fifteen of the 214 patients died (7%). Eight patients died due to complications of

aganglionosis (including 3 patients with total intestinal aganglionosis), and 6 patients died due

to associated malformations. Among these 6 patients were 2 patients with Down syndrome.

who died due to cardiac failure. In one case the cause of death was not known (Table Il),

These figures indicate that the presence of associated malformations is a significant cause of

death.

Short-segment HSCR

In the majority of patients SS-HSCR was the only malformation (n = 131; 77.1 %). Associated

malformations were present in 39 patients (22.9%). We divided these 39 patients into four

additional classes according to the nature of the associated malformation(s). The incidence

~male

o female

length of aganglionic segment

Figure 1: Length of the aganglionic segments. Diagram depicting the percentages of cases

with classic Hirschsprung 's disease (HSCR). long segment aganglionosis (LSA). total colonic,

and total intestinal aganglionosis (TeA and TIA) and the sex. distribution within each class.

124

Table II: Causes of death in patients with intestinal aganglionosis

Death related to the intestinal aganglionosis:

Patient Sex Cause

A41 M Enterocolitis, sepsis

Al32 M Suture leakage, sepsis, multiple organ failure

Al31 M Postoperative bowel perforation, sepsis

Al7 M Suture leakage, sepsis

A45 F Necrotic bowel, septic shock

B32 M Total intestinal aganglionosis

B5 M Total intestinal aganglionosis

A23 F Total intestinal aganglionosis

Death caused by associated malformations

Patient Sex Cause

A49 M Respiratory tract infection, atrio-ventriculo septal defect, mitral valve

insuffiency, aortic coarctation, Down's syndrome

B37 M Complications during surgery of atrio-ventriculo septal defect. Down's

syndrome

A51 M Blalock shunt dysfunction in patient with atrio-ventriculo septal defect.

atresia of tricuspid valve

B54 F Sepsis in patient with multiple congenital malformations

B45 F Disseminated neuroblastoma, Ondine's curse

B42 F Sepsis. ventilatory complications, Ondine's curse

Patient Sex Cause

AlSO M Unknown

and sex ratios of the various classes of SS-HSCR are summarized in Table ill. Class II entails

13 cases with Down syndrome (7.6%). Of these 11 were male and 2 were female. Eight of

these 13 patients also had cardiac malformations. All other SS-HSCR patients with associated

anomalies are presented in Table IV. Class ill entails 10 other 'syndromic' cases (5.9%; d'

125

to ~ .. I to I). These patients had both major and ntinor malformations and craniofacial

dysmorphisms. A syndrome diagnosis was generally not made in these patients. Patient B42

had Ondine's curse (OMIM 209880). Two patients (A152 and A3) had mental retardation and

epilepsy. Limb abnormalities were present in patients AI8 and A48. Three patients (A48,

A24. and A135) had congenital deafness. Class IV entails 4 patients with craniofacial

dysmorphisms only (2.3%; d' to ~, 3 to I). Facial dysmorphism included broad nasal bridge,

antimongoloid eyes. and dysplastic and low set ears.

Class V entails 12 patients with one other major/minor malformation in addition to

HSCR (7.1%; d' to ~ 5 to I). Patient B35 had a pulmonary artery stenosis. Five patients had

a Meckel's diverticulum. Patient B49 also had a Meckel's diverticulum. but we considered

this as part of a 'syndrome', so this patient was ranked class III. Parjent B53 had polydactyly

and rocker bottom feet. Three patients showed vesice-ureteral reflux. of which one also had

hydronephrosis (A72).

We identified 5 families that presented 14 cases with isolated HSCR (d' to ~, 7 : 4;

for the other 3 cases the sex was unknown). All HSCR families are presented in Table V. We

identified 4 families in which SS-HSCR was associated with other malformations. These

families provided 6 cases of SS-HSCR, one case with LS-HSCR and two cases with unknown

Table III:

Class

I

II

III

IV

V

126

Incidence and sex-ratio of the various classes of both short- and long­

segment HSCR

Short-segment HSCR Long-segment HSCR

n % d':~ n % d':~

Total 170 4:1 44 3:1

No associated 131 77.1 4.5:1 24 54.5 7:1

anomalies

Down syndrome 13 7.6 5.5:1 2.2

'Syndromic' 10 5.9 I: I 9 20.5 5:4

Associated with 4 2.3 3: I 5 11.4 2:3

cranio-facial

anomalies

Associated with 12 7.1 5:1 5 11.4 4:1

other anomalies

length of the aganglionic segment (cf to ~. 8 to 1). In proband B50, SS-HSCR was associated

with craniofacial dysmorphisms and nystagmus. A brother of this patient's father had isolated

HSCR. In patient A114, SS-HSCR was associated with Meckel's diverticulum; the mother's

brother had isolated HSCR. In one family SS-HSCR occurred as an isolated malformation in

the proband (A134) and in association with congenital deafness in the brother of this patient

(A135). In one family, a girl with SS-HSCR (AIS) had a brother with LS-HSCR (AI7). Apart

from SS-HSCR. the proband also had craniofacial dysmorphisms, syndactyly and

brachydactyly, and was mentally retarded. Her brother, A17, had similar craniofacial

dysmorphisms. while another brother shared the craniofacial dysmorphisms and the mental

retardation but did not have HSCR.

The family history of 5 SS-HSCR patients without associated malformations revealed

10 family members with several congenital malformations other than HSCR. In the family

of patient Bl. cardiac septal and neural tube defects occurred. In the family of patient A79,

three cases with (cheilo)(gnatho) palatoschisis were present. Pigment abnormalities were

present in family members of patients B21 and A48.

Long-segment HSCR

We ascertained 44 LS-HSCR patients of whom 20 had associated malformations and/or

dysmorphisms (45.5%). We ascertained one male LS-HSCR patient with Down syndrome.

The length of the aganglionic segment in this patient was ca. 20 cm. Nine other 'syndromic'

cases of LS-HSCR were identified (20.5%; d' to ~, I to I). This class (III) included one

female patient (A42) with a mosaicism of an abnormal chromosome 11. The latter patient was

characterized by psychomotor retardation, palatoschisis, webbed neck, hypoplastic nails and

an ectopiC anus. The' syndromic' LS-HSCR patients had both major/minor malformations and

craniofacial dysmorphisms (Table VI). Waardenburg syndrome type 2 was present in cases

A25 and B5~ Ondine's curse and neuroblastoma was present in patient B45. In the other 5

cases a syndrome diagnosis was not made. Class IV patients had associated craniofacial

dysmorphisms (11.4%; d' to ~, I to I). Facial dysmorphisms included hypertelorism,

epicanthus, low set and tilted ears, bifid earlobe, broad nasal bridge, small nose and mouth,

short philtrum and micrognathia. Class V entails 5 patients with one other congenital

malformation (11.4%; d' to ~,4 to I). Three patients (B52, A56, and A23) had abnormalities

of the urogenital tract (hypoplastic or aplastic right kidney). and another had an eventration

of the left diaphragm. Patient A51 had a atrio-ventricular septal defect and tricuspid valve

atresia.

Familial occurrence of LS-HSCR and/or associated anomalies is presented in Table

VII. We identified two families with isolated LS-HSCR which presented 7 patients with

aganglionosis (5d'~ the sex of the other two cases was unknown). In one family two boys had

LS-HSCR. In another family, LS-HSCR occurred in the proband while aganglionosis of

127

Table IV: l\'Ialformations and/or dyslllorphisms associated with SS·HSCR

Class Pat. Sex Head eNS Eyes Ears !l.louthlnose Hem Olher

III 842 F+ Dndine's Sacral dimple

curse

III B27 M Narrow eyelids Low set ear Tenlmouth VSD Eartag left l\licrognathia

III AJ r Asymmetric Retardation Hypertelorism Pylorus hypertrophy

face Epilepsy

III A152 M Retardation Bat ears Broad nasal

Epilepsy bridge

III B50 M Sloping Nystagmus Asymmetric Large ears Large nose Pulm. art. Sacral dimple

forehead eyelids Thin lips stenosis

Deep-set eyes

III B49 M Tilted ears Prominent Cutaneous syndactyly

philtrum and Meckel's diverticulum

upper lip;

l'oticrognathia

III AI8 F l\licrocephaly Blue sclera Abnormal Broad nasal Brachydactyly of hig toes

ea" bridge Syndactyly of dig. 2-3 both

fcel

III A48 F Congenital Adacty!y of dig. 2-4 both feet

deafness

111 A24 F Deafness

111 A13S M Deafness

IV A53 M Epicanthal Broad nasal

folds bridge

IV BI7 M Antimongoioid

eyes

IV B20 F Dysplastic Micrognathia low set ears

IV B58 M Ear tag

V 835 M Pulm. art.

stenosis

V A114 M r-.leckel's diverticulum

V A78 M Meckel's diverticulum

V AlO2 M "'[eckel's diverticulum

V A75 M Meckel's diverticulum

V BI2 M Meckel's diverticulum;

Sacral dimple

V B40 M Single umbilical artery

V B53 M Polydactyly; Rockerbottom

feet

V An F Hydronephrosis; Hydro-

ureters; reflux

V B43 F Hemihypertrophy R

V AI9 M Vesica-ureteral reflux L

V A33 M Vesica-ureteral reflux L

Table V: Familial occurrence of SS-HSCR and/or associated anomalies

Pat. Sex Associated anomalies F:unilial aganglionosis Associated anomalies in family members

A89 M Father

A90 M 3 cases in family

BI6 M Father. 2 sisters of father

AI64 M Brother of A165

AI65 F Sister of AI64

B4 F Mother's father

B50 M Craniofacial Father's brother dysmorphisms Nystagmus

A114 M Meckel's diverticulum Mother's brother

A134 M Brother of A135

A13S M De:rlness Brother of A134

AI8 F Retardation Syndactyly Sister of A 17 Brother: dysmorphisms. Craniofacial retardation dysmorphisms

BI M Sib: cardiac septum defect: Father's sister: cardiac valve anomaly; Mother's sister: Child with Down syndrome. and child with neural tube defect

B21 M Father: white forelock

A79 M Brother. father and brother of father: (cheilo)(gnatho)~ palatoschisis

AlSO M Sib: spina bifida

A48 F Mother: polychromatic iris

unknown length occurred in his father. 2 sibs of the father. and father's uncle. Associated

malformations and/or dysmorphisms were not reported in these families. LS-HSCR was

associated with other congenital malformations in 2 families. In one family Waardenburg

syndrome type 2 was diagnosed (A25 and B5). In the other family. two sibs of the proband

had Potter's syndrome.

The incidence of associated anomalies and sex ratios of the various classes of SS- and

LS-HSCR are summarized in Table ill. Statistical analysis revealed that 1) There is no

130

significant difference between SS-HSCR and LS-HSCR with regard to sex ratio (p ; 0.605).

2) LS-HSCR is more frequently associated with anomalies than SS-HSCR (p < 0.005); 3)

Down syndrome is most frequently associated with SS-HSCR (p < 0.05). 4) There are

significant differences with regard to sex-ratios in class ill and IV compared to class I (p <

0.0001).

Discussion

As an initial step to study possible pathogenetic mechanisms of one or more types of HSCR.

we detennined the incidence and nature of associated malformations and/or dysmorphisms in

patients. Establishing the occurrence of associated malformations might help in syndrome

delineation.

Retrospective analysis of 214 HSCR patients revealed an incidence of 27.6% of

associated malformations and/or dysmorphisms. This percentage is much higher than that

found in the normal population (2.3 to 5% of liveborn neonates [Mohes. 1988. Reerink et al .•

1993 J. The incidence of congenital malformations associated with HSCR reported in the last

15 years ranges from 5.3% to 29.8% [Suzuki et al .• 1978. Ikeda and Goto. 1984. Spouge and

Baird. 1985. Moore et al .• 1991. Ryan et al .. 1992]. In two recent surveys of 370 and 179

HSCR cases respectively, the incidence of associated malformations amounted to 16.5% and

22% [Moore et al .. 1991. Ryan et al., 1992]. It is likely that associated malformations have

been underreported in the past, both in our institute and in the literature. This becomes

particularly evident when we compare the incidence of associated malformations in our A

group. admitted between 1970 and 1985 (19.8%) and B group. admitted between 1985-1992

(48.3%). It seems that routinyevaluation by a dysmorphologist is an important factor in the

detection of associated anom,Dies. When we considered SS-HSCR and LS-HSCR separately

the incidence of associated anomalies amounted to 22.9% and 45.5% respectively. Ikeda and

Goto [1984] also reported that the incidence of associated malformations increased with the

length of the aganglionic segment (SS-HSCR 10.2%; total colonic aganglionosis 15.2%). They

did not investigate the nature of the associated malformations.

Pattern of malformations and .dysmorphisms associated with HSCR

In the majority of SS-HSCR and LS-HSCR cases in this study, aganglionosis seems to be an

isolated defect (77.1 % and 54.5% respectively). In 22.9% and 45.5% of cases. SS- and LS­

HSCR occurred in association with one or more congenital malformation. We classified these

remaining cases depending on the phenotype of the associated malformations. We found that

SS-HSCR and LS-HSCR do not differ markedly with respect to the pattern of associated

anomalies. sex ratio, and familial occurrence. It might well be that isolated forms of short-

131

Table VI: Malformations andlor dysmorphisms associated with LS-HSCR

Class Pat. Sex Head eNS Eyes Ears Mouthlnose Heart Other

III A 103 M Flattened Retardation Epicanthus Low set ears Large mouth Hypospadia; Partial

forehead Epilepsy Horner L Micrognathia syndact. dig. 2-3 L hand

III Bll M Protruding Epicanthus L Broad nasal Truncus Abnormal footcreases

forehead bridge arteriosus

III A42 F Retardation Palatoschisis Webbed neck; Ectopic

anus (chrom 11 abo.)

III B54 F+ Round head Hypertelorism Dysplastic Rear rvlicrognathia Pulm. art. Dysplastic kidneys

Microcephaly Iris coloboma Deafness stenosis

III D30 M Hypertelorism Hypoplastic R kidney

Widow's peak;

Syndactyly

III B33 M Micrognathia Pulm. art. Syndactyly

stenosis

III A25 F Pale irides Dysplastic ears; Broad nasal

Congenital bridge

deafness

III B45 F+ Ondioe's curse Neuroblastoma

III B5 M+ Pale jrides Dysplastic ears White hairlock; Sacral

dimple; Lumbal patch of

hair

IV AI7 M+ Short neck Low sct ears Broad nasal

bridge

IV B4I F Micrognathia

IV B31 F Round head Hypertelorism Small nose

Sloping Small mouth

forehead

IV A7 F Bifid earlobe L

IV B32 M+ Epicanthus Tilted ears Short Abnormal hand crease R philtrum Micrognathia

V B52 M Vesica-ureteral reflux R

V AS M Eventration of

diaphragm L

V A56 M Hypoplastic R kidney

V A5I M+ AVSD;

tricuspid atresia

V A23 F+ Abnormal handcreases Aplastic R kidney

and long segment disease are variant forms of the same genetic defect.

The association of HSCR and Down syndrome has since long been recognized [Wolf

and ZweymuelIer, 1962, Emanuel et aL 1965]. The incidence of Down syndrome in our study

(6.5%) falls within the 2% - 14% range reported in the literature [Passarge, 1967, Polley and

Coran, 1986]. It is striking that 13 of the 14 cases had SS-HSCR. This is in accordance with

the study by Badner and coworkers [1990]. In addition, the male to female preponderance of

6 : 1 in our series was also noticed by this group (10.5:1). We were not able to determine

whether this male preponderance in the combination of SS-HSCR and Down syndrome also

occurred in other large series since the sex of the Down cases is rarely mentioned. Trisomy

21 is a major cause of congenital malformations of the heart and digestive tract and of mental

retardation. Recent work has suggested small regions in band q22 that are likely to contain

genes for some of these features [Korenberg et al .. 1992J. However. until now no linkage

between this region and HSCR has been reported [Slaugenhaupt et al., 1991]. Various

nonspecific mechanisms such as increased cell adhesion and limited cell proliferation. have

been suggested as pathogenetic mechanisms for the malformations. but the relationship of

these cellular events to chromosome 21 imbalance is not understood.

In total. 19 'syndromic' cases were recorded, representing 5.9% of the SS-HSCR cases

and 20.5% of the LS-HSCR. Some of these malformations and dysmorphisms associated with

HSCR are similar to known Mendelian disorders or partially resemble them. A syndrome

diagnosis was obtained only in 5 cases (two Ondines curse (OMIM 209880), three cases with

Waardenburg syndrome type 2). The association of HSCR and Waardenburg syndrome type

I is the best known [Omenn and McKusick, 1979, Badner and Chakravarti, 1990].

Waardenburg syndrome manifests as variable combinations of deafness and pigment

abnormalities such as heterochromia iridis. white forelock and white skin patches. Clinically,

Waardenburg syndrome is divided into two main subtypes which usually breed true within

families and which have been supposed to be genetically distinct. The two types are

distinguished by the presence in type 1 of dystopia canthorum. an outward displacement of

the inner canthi of the eyes. Waardenburg type 1 has been attributed to mutations in the PAX3

gene. There is no consensus whether mutations in PAX3 can also lead to Waardenburg

syndrome type 2. Until now no linkage between the PAX3 region on chromosome 2 and

HSCR has been reported [Slaugenhaupt et al., 1991]. Waardenburg syndrome type 2 was

found to be associated with both SS-HSCR and LS-HSCR within one family (patients A25

and B5). We identified five families in which the proband suffered from isolated SS-HSCR

while other family members had features occurring in the Waardenburg syndrome (white

forelock, polychromatic iris, spina bifida). It might well be that the association of HSCR and

Waardenburg syndrome type 2 is an autosomal dominant trait with variable penetrance and

expression. Detailed family histories of sporadic HSCR cases might help to elucidate this

issue.

134

Table Vll: Familial occurrence of LS-HSCR audlor associated anomalies

Pat. Sex Associated Familial Associated anomalies anomalies aganglionosis in family

B24 M Father, 2 sibs of

father, father's uncle

A20 M Brother of A21

A21 M Brother of A20

AI7 M Craniofacial Brother of AI8

dysmorphisms

A25 F Waardenburg type 2 Sister of B5

B5 M Ear malformations Brother of A25

Piebaldism

Micrognathia

B32 M Craniofacial and other Two sibs: Potter

dysmorphisms syndrome

A clear 'syndrome' diagnosis was not made for the sibs A18, A17, One brother had

similar dysmorphisms but no HSCR. These patients resemble the cases described by Hurst

et aI. [1988]. HSCR patients without overt congenital malformations but with craniofacial

dysmorphisms might also be included in the 'syndromic' cases of HSCR.

In 17 patients (7.9%), HSCR was associated with one other congenital malformation

but without craniofacial dysmorphisms. We found Meckel's diverticulum in 3.8% of patients.

compared to 2% in the general population [Turgeon and Barnett, 1990, St-Vil et aI., 1991],

and to 0.5% in another group of HSCR patients [Ikeda and Goto, 1986]. Cardiac defects and

single umbilical artery have been observed as non-random occurring associated anomalies.

These isolated congenital malformations can occur together in the VATER or V ACTERL

association. A co-occurrence of HSCR and a full blown V ACTERL association has been

reported by Ryan et aI. [1992]. In other studies association of HSCR with either anal atresia,

limb, cardiac, or urinary tract malformations have been reported [Spouge and Baird. 1985.

Takada et aI., 1985, Ikeda and Goto, 1986. Watanatittan et aI., 1991].

We classified HSCR into at least four different classes based on the presence and

nature of associated malformations. The developmental biology of the enteric nervous system

might help in elucidating the pathogenesis of the various forms of HSCR. The neurons and

135

supportive cells of the vertebrate enteric nervous system derive from neural crest cells

originating from the posterior hindbrain. During an early phase of development. the hindbrain

is divided into eight repeating units or rhombomeres [Keynes and Lumsden, 1990, Lumsden

et aI.. 1991, Noden. 1991]. In the anterior hindbrain. the neural crest is segregated into

streams. It is unknown whether neural crest cells in the posterior hindbrain are also segregated

into streams. Ablation of the neural crest adjacent to somites 3 through 5 resulted in

aganglionic colon without any other obvious malformations. Ablation of the neural crest of

rhombomeres 6 through 8 resulted in total intestinal aganglionosis. thymic aplasia or

hypoplasia. and disturbances in the cardiac outflow tract [Kirby. 1993. Peters-van der Sanden

et al .. 1993]. Apparently. the longer the aganglionic segment the higher the incidence of

associated malformations. Based on these experimental data one would predict that the

malformations associated with HSCR reside in derivatives of the neural crest of the posterior

hindbrain (or third and fourth pharyngeal arches). This holds true for the 7 cases with cardiac

outflow defects in our series and cases with cardiac defects (e.g. tetralogy of Fallot) in other

studies [Spouge and Baird. 1985. Ryan et al .. 1992]. This might also hold true for

abnormalities in the thyroid and parathyroids as these occur in the MEN-2A syndrome [Verdy

et al .. 1982}. However, these abnormalities consist of tumors instead of congenital

malformations. An association with thymic hypo-or aplasia has not been reported.

Craniofacial dysmorphisms and malformations can be related to the cranial neural crest

since this crest contributes to the ectomesenchymal structures in nasal, orbital. maxillary

skeleton, palate. cranial bones, and the otic capsule. as well as to connective tissue of the face

and ventral part of the neck [Couly et al .. 1993].

Genetic factors and HSCR

Several lines of evidence suggest the involvement of genetic factors in the pathogenesis of

HSCR i) the elevated risk to sibs: ii) the dominant pattern of inheritance in several families

with LS-HSCR: iii) the association with trisomy 21, microdeletions of chromosome 13q, and

10q: iv) the presence of linkage of LS-HSCR with a locus on chromosome 10qll: and v) the

existence of Mendelian models for colonic aganglionosis in rodents. The pattern of

inheritance, however. does not appear to be due to a single gene in all families.

We identified 8 familial cases of isolated HSCR (4.2% of all cases). There was no

apparent difference with regard to the incidence of familial cases between familial SS- and

LS-HSCR cases. When 'syndromic' cases are included, however, familial occurrence

increased with the length of the aganglionic part of the bowel (6.5% vs 14%: the overall

incidence of familial cases is 7.9%). This has been observed previously [passarge. 1967.

Badner et al .• 1990. Moore et al .. 1991].

The majority of our HSCR patients were not karyotyped. There are only few examples

in the recent literature concerning cytogenetic abnormalities in patients with intestinal

136

aganglionosis. Of particular interest is the observation by Martucciello et al. [1992] of a small

cytogenetically visible deletion of chromosome IOq11.2-q21.2 in a female patient with LS­

HSCR without any other detectable anomalies. This particular patient led to the mapping of

familial LS-HSCR to a locus on the proximal long arm of chromosome 10 (lOqI1.2) between

DIOS20S and DlOSI96 [Angrist et aI., 1993, Lyonnet et aI., 1993]. Edery et aI. (submitted

for publication) described the c-RET proto-oncogene as the closest genetic marker with

respect to the disease locus, suggesting that this proto-oncogene, which has been shown to

account for multiple endocrine neoplasia type 2A, might be a candidate gene for HSCR. For

both familial LS- and SS-HSCR, tight pairwise linkage with no recombination events was

observed between c-RET and the disease loci (personal communication). suggesting that

familial SS-HSCR and LS-HSCR are allelic disorders. It might well be that chromosome

109 11 contains a set of genes that are important for the development of the rhombencephalic

neural crest.

Acknowledgements

We thank Dr. Richard Langemeijer for the use of his HSCR database; we also thank the

members of the genetic counselling team (Department of Clinical Genetics, Academic

Hospital Dijkzigt and Erasmus university) for their comments on the patient material, Dr. A.P.

Provoost for aid in statistical analysis, and Mr. Ko Hagoort for editorial assistance.

References

Angrist M, Kauffman E, Slaugcnhaupt A, Matise TC, Puffenberger EG, Washington SS, Lipson A, Cass DT, Reyna T, Weeks DE, Sieber W, Chakravarti A (1993): A gene for Hirsehsprung disease (megacolon) in the pericentromcric region of human chromosome 10, Nature Genetics 4:351~356.

Badner JA, Chakravarti A (1990): Waardenburg syndrome and Hirschsprung disease: evidence for pleiotropic effects of a single dominant gene. Am. J. Med. Genet. 35:100~104,

Badner JA. Sicber WK. Garver KL, Chakravarti A (1990): A genetic study of Hirschsprung disease. Am. 1. Hum. Genet. 46:568~580.

Couly GF, Coltey PM, LcDouarin NM (1993): The triple origin of skull in higher vertebrates: a study in quail~chick chimeras. Development 117:409~429.

Ehrenpreis T (1970): "Hirschsprung's disease", Chicago.

Emanuel B, Padorr MP. Swenson 0 (1965): Mongolism associated with Hirschsprung's disease. J. Pediatr. 66:437.

Hurst JA, Markiewicz M, Kumar D, Brett EM (1988): Unknown syndrome: Hirschsprung's disease, microcephaly, and iris coloboma: a new syndrome of defective neuronal migration. J, Med. Genet. 25:494-500.

Ikeda K. Goto S (1984): Diagnosis and treatment of Hirschsprung's disease in Japan. Ann. Surg. 199:400~405.

Ikeda K. Goto S (1986): Additional anomalies in Hirschsprung's disease: an analysis based on the nationwide survey in Japan. Z. Kinderchir. 41:279~281.

Keynes R, Lumsden A (1990): Segmentation and the origin of regional diversity in the vertebrate central nervous system, Neuron 4:1~9.

Kirby ML (1993): Cellular and molecular contributions of the cardiac neural crest to cardiovascular development. Trends in Cardiovascular Medicine 3:18~23.

137

Kleinhaus S. Boley S1. Sheran M. Siebcr WK (1979): Hirschsprung's disc:r.se. A survey of the members of the surgical section of American academy of pediatrics. 1. Pediatr. Surg. 14:588-597.

Korenberg JR, Bradley C. Disteehe eM (1992): Down syndrome: molecular mapping of the congenital heart disease and duodenal stenosis. Am. J, Hum. Genet. 50:294-302.

Lumsden A. Sprawson N. Graham A (1991): Segmental origin and migration of neural crest cells in the hindbrain region of the chick embryo. Development 113:1281-1291.

Lyonnet S, Bolino A, Pclet A. Abel L, F6kete CN. Briard ML. Siu VM, Kaariainen H, Martueeiello G, Lcrone M. Puliti A, Luo Y. Wcissenbaeh J, Devoto M. Munnich A. Romeo G (1993): A gene for Hirsehsprung disease maps to the proximal long :lI'Ill of chromosome 10. Nature Genetics 4:346-350.

Martueeiello G. Bicocehi MP, Dodero p, Lcrone M. Cirillo MS. Puliti A. GimeIli G, Romeo G, Jasonni V (1992): Total colonic aganglionosis associated with interstitial deletion of the long arm of chromosome 10. Pediatr. Surg. Int. 7:308-310.

M6hcs K (1988): "Informative morphogenetic variants in the newborn infant", Budapest Akad6miai Kiad6.

Moore SW, Rode H. Millar AJW, Albertyn R. Cywes S (1991): Familial aspeets of Hirschsprung's disease. Bur. J. Pediatr. Surg. 1:97-101.

Noden DM (1991): Vertebrate craniofacial development: The relation between ontogenetic process and morphological outcome. Brain Behav. Evo!. 38:190-225.

Omenn GS. McKusick VA (1979): The association ofWaardenburg syndrome and Hirschsprung megacolon. Am. J. Med. Genet. 3:217-223.

Passarge E (1967): The genetics of Hirschsprung's disease. Evidence for heterogeneous etiology and a study of sixty-three families. New Eng!. J. Med. 276:138-143.

Peters-van der Sanden MJH. Kirby ML. Gittcnberger-de Groot A. Tibbocl D. Mulder MP, Meijer:; C (1993): Ablation of various regions within the avian neural crest has differential effects on ganglion formation in the fore-, mid-, and hindgut. Dev, Dyn. 196:183-194.

POlley JTZ. Coran AG (1986): Hirsehsprung's disea..~e in the newborn. Pediatr. Surg. Int. 1:80-83.

Reerink JD. Herngreen wp, Verkerk PH, Ruys JH. Verloove-Vanhorick SP (1993): Congenitalc afwijkingen in het eerste lcvensjaar in Nederland. Ned. Tijdsehr. Geneeskd. 137:504-509.

Ryan ET, Ecker JL, Christakis NA. Folkman J (1992): Hirsehsprung's diSC-aSe: associated abnormalities and demography. J. Pediatr. Surg. 27:76-81.

Slaugcnhaupt SA. Angrist M, Leariseh RD. Lcsh DB. Washington SS. Chakravarti A (1991): Hirsehsprung disease: linkage analysis of candidate regions on human chromosome 2. 13, and 21. Am. 1. Hum. Genet. 49:359,

Spouge D. Baird PA (1985): Hirschsprung's disease in a large birth cohort. Teratology 32:171-177.

St-Vil D. Brandt ML. Panic S. Bensoussan AL, Blanchard H (1991): Meckel's diverticulum in children: a 20-year rcview. J. Pediatr. Surg. 26:1289-1292.

Suzuki H. Chibil T. Kasai M (1978): Entire colon aganglionosis and extensive ilganglionosis: an analysis of 94 cases in Japan. Jap. J. Surg. 8:119-122.

Takada Y. Aoyama K. Goto T, Mori S (1985): The association of imperforatc anus and Hirschsprung's disea..~ in siblings. J. Pediatr. Surg. 20:271-273.

Turgeon DK. Barnett JL'(1990): Meckel's diverticulum, Am. J. Gastroentcrol. 85:777-781.

Vcrdy M. Weber AM. Royer CC. Morin CL. Cadotte M. Brochu P (1982): Hirsehsprung's discase in a family with mUltiple endocrine neoplasia type 2. J. Pediatr. Gastroenterol. Nutr. 1 :603-607.

Watanatittan S, Suwatanaviroj A, Limprutithum T. Rattanasuwan T (1991): Association of Hirsehsprung's disease and anorectal malformation. J. Pediatr. Surg. 26:192-195.

Wolf HG. Zweymueller E (1962): Mongolism and aganglionic megacolon. Wcin. Klin. Wehnsehr. 74:219.

138

3.8. General discussion

Finally, I would like to try and make an inventory of the certainties and, perhaps even more

importantly. the uncertainties, of ENS development. Because of its complexity, I subdivided

this process into a number of separate components. although onc should realize that such a subdivision is somewhat artificial.

1. Segmentation within the neural crest regarding enteric nervous system development

In recent years. research interest has become focused on segmentation of the neural crest.

especially in the region of the rhombencephalon. It was shown that within this region. small

neural crest segments gave rise to specific derivatives (Naden. 1983: Lumsden et al .. 1991).

In the posterior rhombencephalon. ectomesenchymal derivatives originate mainly from the

cardiac neural crest from the level of the mid~otic vesicle down to the caudal boundary of

somite 3 (rhombomeres 6-7) (Kirby et al .. 1983: Bockman and Kirby, 1984). Inversion of this

part of the neural crest resulted in defective cardiac development, further indicating its segmental character (Kirby, personal communication). A major finding in this thesis is, that

the neural crest adjacent to somites 3-5 (rhombomere 8, and perhaps part of rhombomerc 7)

is specifically responsible for the innervation of the hindgut (colon) in vivo. Construction of

isotopic and isochronic quail-chick chimeras showed that this small segment of the posterior

rhombencephalic neural crest could even be responsible for the innervation of the entire digestive tract (LeDouarin and Teillet. 1973) (Peters-van der Sanden et al .. unpublished

results). We sunnise that the posterior rhombencephalic neural crest is segmented with regard

to the formation of the ENS. In vivo. the innervation of the hindgut. and possibly of the entire gut, specifically depends on the neural crest adjacent to somites 4-5.

The molecular basis for the segmentation of the rhombencephalic crest could be

formed by the differential antero-posterior expression of members of the homeobox-containing gene family. In mice. study of the expression pattern of these genes in both the

neuroepithelium and the paraxial mesoderm has shown that each axial segment expresses a

specific combination of genes, providing it with a so-called Box code (Hunt and Krumlauf,

1992). Analysis of transgenic mice bearing targeted germline mutations in members of the

Box-A gene cluster has provided direct experimental evidence that these Box genes play an important role in hindbrain segmentation. A knockout-mutation of the Box-AI gene affected

the development of cranial ganglia, whereas a knockout mutation of Box-A3 mainly affected

ectomesenchymal derivatives of the rhombencephalic neural crest (Chisaka and Capecchi,

1991: Lufkin et al .. 1991: Chisaka et al .. 1992). Overexpression of the Hox-A7 gene resulted

in defects in the cervical vertebrae and in craniofacial abnormalities (Balling et al .. 1989: Kessel et al .. 1990). whereas overexpression of the Box-A4 gene resulted in disturbed ENS

139

~

No neural crest is produced at the levels of r3 and r5

Neural crest from mesencephalon and cranial end of metencephalon migrates to arch 1

Neural crest from superior end of myelencephalon migrates io arch 2

A-O: Vagal neural crest

[

A: migrates to arch 3

Cardiac crest B: migrates to arch 4

C: migrates to arch 6

Enteric crest 0: does not migrate through the arches

I ;::~ ~~-

~

~ ~

~

.......

r1

s

2

3

4

5

0 ,... (\J Cl) >< >< ~ 0

::t:

T L[ r

(\J Cl)

~

.J ...... _0'

<') '" Cl Cl ('oj' ",-Cl) Cl)

<')- ",- '" « « () >< >< >< ~ 0 ~ ::t:

Figure 11: Schematic presentation oflhe variolls regions afthe Ileural crest ami their relation to the rhombomeres lind phm)'Jlgea/ arches. Segmentation within tile neural crest is correlated with Hox gene e;rpressioll in the neuroepithelium.

development (Wolgemuth et al., 1989). The fact that Hox gene expression can be altered by

RA (Marshall et al., 1992), could provide the molecular mechanism that underlies the

selective developmental defects seen in RA embryopathy (chapter 3.3).

In Fig. 11. segmentation of the posterior rhombencephalic crest with regard to the

formation of its various derivatives, is correlated with the expression patterns of various Hox

genes in mouse embryos. We surmise that the posterior rhombencephalic neural crest can be

subdivided into the cardiac crest, from the midotic vesicle down to the caudal boundary of

somite 3, and the enteric crest adjacent to somites 4-5. Detailed analysis of the expression

patterns of the members of the various paralogous groups in both chicken and mouse embryos

could provide further insight into the extent of segmentation of the rhombencephalic neural

crest.

2. Migration of neural crest cells to the gut

In order to be able to contribute to ENS development, neural crest cells have to leave the

neural tube, migrate to the gut and enter it at some point. There are still a number of largely

unanswered questions, regarding these processes. First of all, when do neural crest cells.

especially those at the level of somites 4-5. leave the neural tube? Studies using quail-chick

chimeras have shown that most precursors for enteric neurons have left the neural folds prior

to the 13-somite stage (HH stage II) (LeDouarin and Teillet, 1973). Newgreen and Erickson

(1986) showed that neural crest cell migration at this level occurred between somite stages

10-22 (RR stage 10-14). We showed that neural anlagen, consisting of neural tube, notochord

and small somitic remnants. obtained from embryos having more than 18 somites (HH stage

13) were better capable of forming enteric ganglia than neural anlagen obtained from younger

embryos (chapter 3.2.). These results might indicate th2l.t the enteric precursors either leave

the neural crest later than hitherto assumed, or that they remain in close contact with the

neural tube for a prolonged period of time. This delayed migration could represent some kind

of induction or maturation process, dependent on close contact with the neural tube or

notochord. necessary for proper ENS formation. Signals emanating from the neural tube were

found to be necessary for the development of dorsal root ganglia in the trunk region

(Kalcheim and LeDouarin, 1986; Kalcheim et al .. 1987). Differentiation into enteric neurons.

however, can occur in the absence of the notochord (Teillet et al., 1978).

Our second question concerns the migration pathways taken by the enteric neural crest

precursors on their way to the gut. Recently, the migration pathways of rhombencephalic

neural crest cells have been studied in great detail (Kuratani and Kirby. 1991: Lumsden et al .•

1991; Miyagawa-Tomita et al., 1991; Sechrist et al .. 1993). These studies have shown that

cells from the rhombencephalic neural crest dovm to the caudal boundary of somite 3 migrate

predominantly via a dorsolateral pathway on their way to the pharyngeal arches. whereas

141

caudal to somite 3, neural crest cells no longer enter the pharyngeal arches and instead

migrate along a ventrolateral pathway through the anterior part of the somites. In this thesis,

we show that the neural crest caudal to somite 3, is primarily responsible for ENS formation

at least in the colon, indicating that ENS precursors might migrate along a ventrolateral

pathway on their way to the gut. In vitro clonal analysis of the posterior rhombencephalic

neural crest showed that the potential to differentiate into serotonergic neurons, which may

constitute precursors for enteric neurons, decreased considerably upon entry into the

pharyngeal arches (Ito and Sieber-Blum, 1993). Preliminary results, using microinjection of

the lyophilic dye DiI, labelling a specific segment of the neural crest, further support the

hypothesis that precursors for enteric neurons do not migrate through the pharyngeal arches

(Raams et al., unpublished results). Previous studies that suggested that the enteric precursors

migrate along a dorsolateral pathway, through the caudal pharyngeal arches (Thiery et al.,

1982; Ciment and Weston, 1983; Payette et al., 1984; Tucker er al., 1986), were primarily

based on indirect evidence provided by immunohistochemical localization of presumptive

enteric precursors in tissue sections, using either an anti-neurofilament or the HNK-l (NCl)

antibody (Thiery er al., 1982; Payette et al., 1984; Tucker er al., 1986). These results provide

only a static picture and do not show that cells, immunoreactive for a certain antibody,

observed within the branchial arches, actually migrate to the gut. Additional evidence for a

difference in migration pathways between anterior and posterior vagal neural crest cells comes

from experiments, in which small segments of the vagal neural crest were cocultured with

aneural hindgut on the chorioallantoic membrane. These experiments showed that all segments

of the posterior rhombencephalic neural crest had the developmental potential to colonize the

hindgut, indicating that there are no real intrinsic differences between these segments, and that

the specific need for the neural crest adjacent to somites 3-5 for hindgut innervation might

be due to a difference in migration pathways.

Using chorioallantoic membrane cocultures of chicken aneural and quail neural

hindgut, we found that aneural gut was able to attract vagal neural crest cells which had

already colonized a gut segment, provided ganglion formation in the neural gut was not yet

fully completed (chapter 3.4.). Such a specific interaction between aneural hindgut and vagal

neural crest cells was also found in heterotopic quail-chick chimeras in which the vagal neural

crest was transplanted to the adrenomedullary region (LeDouarin and Teiller, 1974). In these

chimeras, quail cells were found in all normal trunk derivatives, but in addition to these sites,

they were also present in the postumbilical gut, a site which normally is not colonized by the

adrenomedullary neural crest. These results indicate that vagal neural crest cells are

specifically attracted by aneural gut. probably by signals emanating from the enteric

mesenchyme. It is generally agreed upon, that neural crest cells enter the pharynx at the third day

of embryonic development (RR stage 17-18) (LeDouarin and Teillet, 1973; Tucker et al.,

142

Neural crest cell migration

A 81-3

B 84-5

C 86->

~~~""'.~

,,8

~

~~ ~' c::::)

tf o ~~~~"

C"i0'0!

o Figure 12: Neural crest cell migration pathways at various levels in the vagal neural crest region. In the lefthand part of each picture, the relative importance of the dorsolateral (between ectoderm and dermamyotome) and ventral (through the anterior part a/the somites) pathways is indicated by the relative size of the arrows. In the righthand picture the localization of the neural crest cells is indicated by the hatched areas. A) at the level of somites 1-3 B) at the level of somites 4-5 C) caudal to somite 6. NT= neural tube; No= notochord; DA= dorsal aorta; DM= dermamyotome; 51= sclerotome; Ph= pharynx; Pp= pharyngeal pouch; CC= circumpharyngeaZ crest; Fg= foregut; H= heart; Ee= enteric crest; G= gut; DRG= dorsal root ganglion; SG= sympathetic ganglion.

143

1986), but the exact site of entry is hitherto unknown. Immunohistochemical data, using either

the HNK-1 (Raams et al., unpublished results) or the ElC8 (Tucker et al., 1986) antibody,

suggest that the point of entry lies at the level of the fifth or sixth somite, but definite proof

should come from in vivo labelling studies using DiI or retroviral markers.

We surmise that the precursors for enteric neurons, predominantly arising at the level

of somites 4-5. do not enter the pharyngeal arches, but migrate ventrolaterally through the

anterior part of the somites and enter the pharynx at the level of somite 5 or 6 (Fig. 12).

3. Migration of neural crest cells through the gut

Once neural crest cells have entered the gut, they have to move caudally in order to give rise

to enteric ganglia along the entire digestive tract. Whether this is achieved by active migration

or by passive displacement along with the elongation of the gut. is still unclear. There are,

however, some indications that, at least in the hindgut, active migration is taking place

(Tucker et aI., 1986). Study of the time course of neural crest migration through the gut

showed that they colonize the foregut at E4. Subsequently, they migrate through the midgut

at E5, reaching the level of the ceca approximately at E6.5. Migration is completed before

E8 (LeDouarin and Teillet, 1973; Allan and Newgreen. 1980). Sacral neural crest cells have

already colonized Remak's ganglion at E4, but they do not enter the hindgut prior to E7

(LeDouarin and Teillet, 1973; Pomeranz and Gershon. 1990; Pomeranz et al., 1991a;

Serbedzija et al., 1991).

At the time neural crest cells move through the fore- and midgut the various layers

of the gut wall have not yet formed, and the gut consists only of an endoderma1 tube

surrounded by splanchnic mesoderm. By the time neural crest cells reach the hindgut the

smooth muscle layers have formed, and neural crest cells are found on either side of these

(LeDouarin and Teillet. 1973; Tucker et al., 1986). We found that neural crest cells initially

migrate superficially under the splanchnic epithelium, forming the myenteric plexuses.

Subsequently, they transverse the muscle layers to give rise to the submucous plexuses

(Souren et al., unpublished results). In the fore-, and midgut, neural crest cells were found not

to migrate as single cells, but by extending processes to their neighbours. thus form a

complex network in the wall of the gut (Epstein et al .• 1991). We found a similar network in

the neural hindgut after wholemount staining using the HNK-l antibody (Nurmohamed et aI.,

unpublished results). Formation of such a network was found to be dependent on the presence

of vagal neural crest cells and thought to be mediated by the enteric microenvironment.

4. Homing of neural crest cells in the gut

While migrating through the gut. neural crest cells have to receive a signal from the enteric

144

microenvironment, which tells them where to stop migration and form enteric ganglia. Upon

study of the microenvironment of the aneural hindgut, we identified a layer of mesenchymal

cells on the luminal side of the smooth muscle layer and cells at the serosal side, which

reacted with the HNK-I antibody (HNK-I mode I; chapter 3.5. and 3.6.). The localization

of these HNK-l immunoreactive cells exactly at the sites where ganglion formation occurs,

suggests that these cells may present a homing signal to neural crest cells. In neural gut, these

HNK-l immunoreactive mesenchymal cells were no longer present and instead the HNK-l

antibody identified the enteric ganglia (HNK-l mode 2). Using an adhesion assay, in which

we incubated isolated neural crest cells on cryosections of aneural hindgut, we confirmed that

these HNK-I immunoreactive cells provided a homing signal for neural crest cells

(unpublished results). We identified two HNK-I carrying, cell membrane glycoproteins of 42

and 44 kD, but purification of these glycoproteins in sufficient amounts for further

characterization and sequencing proved to be difficult. because of the limited amount of

starting material (T. Luider. personal communication). Maybe the recently introduced

technique of differential display using degenerate primers and peR could provide a better tool

(Liang and Pardee, 1992).

Pomeranz et al. (1991 b) also provided evidence for an important role of the enteric

microenvironment in the homing of neural crest cells. They found that neural crest cells

acquire a receptor for laminin, while migrating through the gut. Laminin is an extracellular

matrix molecule, which is normally present in the basal laminae of the mucosal and serosal

epithelium and of the smooth muscle cells of the gut. The interaction between laminin and

its receptor might cause neural crest cells to stop migrating and aggregate into enteric ganglia.

Overabundance of laminin in a broad zone of the enteric mesenchyme, both in a mutant

mouse strain with congenital megacolon (Payette et al., 1988) and in patients with

Hirschspung disease (Parikh et al .. 1992) also points to a possible role for laminin in ENS

formation. In vitro analysis of cell-matrix interactions showed that the attachment of neural

crest cells to extracellular matrix molecules is mediated predominantly by integrin receptors

(Lallier and Bronner-Fraser, 1991). Furthermore. it was shown that crartial and trunk neural

crest cells use a functionally distinct set of integrins to attach to different conformations of

laminin (Lallier et al., 1992).

5. Differentiation of neural crest cells and the role of the enteric microenvironment Using coculture experiments, we showed that the switch from HNK-I mode 1 into HNK-l

mode 2 only occurred when neural crest cell colonization was followed by differentiation into

enteric neurons (chapter 3.5.). When trunk neural crest cells colonized the hindgut. they

differentiated mainly into melanocytes and the HNK-I mode I immunoreactivity persisted.

This indicates that the HNK-l immunoreactive mesenchymal cells in aneural hindgut not only

145

present a migration and homing signal, but might also constitute a differentiation signal for

neural crest cells. But, whereas the migration and homing signal can be recognized by all

neural crest cells entering the hindgut, the differentiation signal is specific for vagal neural

crest cells. Isolated vagal neural crest cells, which were cultured for one day in vitro, retained

the specific characteristics enabling them to differentiate into enteric neurons in the hindgut.

These characteristics, however, were lost upon prolonged in vitro culture, leading them to

behave as trunk neural crest cells, differentiating into melanocytes.

Aberrant differentiation of trunk neural crest cells, which we observed in our coculture

system, was also observed in a more in vivo situation. Le Douarin and Teillet (1974)

constructed heterotopic quail-chick chimeras in which adrenomedullary neural crest was

transplanted to the vagal region, and found that these quail trunk neural crest cells

differentiated into melanocytes in the postumbilical gut. In the preumbilical gut, however,

trunk neural crest cells were able to differentiate into enteric neurons. These results indicate

that there may be intrinsic differences between the microenvironment of the various segments

of the gut and that embryonic cell types respond differentially and specifically to these

signals. Experimental evidence for such a segmentation within the digestive tract comes from

our ablation experiments. in which we showed that there are differences between the various

parts of the digestive tract regarding their dependence on specific neural crest segments

(chapter 3.2.). Whereas the hindgut depended specifically on the neural crest of somites 3-5,

the midgut could be colonized by the entire posterior rhombencephalic neural crest. The

foregut could also be innervated by a source outside the vagal neural crest. presumably by

more anterior rhombencephalic neural crest. Transgenic mice with a null mutation of the ret-l

gene had total intestinal aganglionosis. but they, too, had a normally innervated foregut (V.

Pachnis, personal communication).

These experimental data strongly suggest that there are qualitative differences in the

microenvironment of the various parts of the digestive tract. A quantitative effect. however,

caused merely by the amount of vagal neural crest cells that eventually reaches the gut. can

not be excluded. Such a quantitative effect, however. would be expected to result in a gradual

transition between the ganglionic and aganglionic parts of the gut. We. however, found sharp

boundaries localized at morphOlogically identifiable sites, one at the level of the foregut­

midgut transition, which is the site were the pancreas develops. and the other at the ileo-cecal

junction. The fact, however. that both the pancreas and the ceca are colonized by neural crest

cells. might suggest that the migrating neural crest cell-network arrests temporarily at these

sites until either the pancreas or the ceca are colonized, thereby increasing the chance that the

transition of ganglionic to aganglionic gut is found at these sites.

We conclude that neural crest cells carry receptors that recognize signals in the enteric

microenvironment. which enable them to migrate through the gut, home at specific sites and

differentiate into enteric neurons. The microenvironment of the postumbilical gut harbors

146

signals, that differ from those in the preumbilical gut, and only vagal neural crest cells are

able to recognize these specific differentiation signals presented by the HNK-l

immunoreactive cells. Trunk neural crest cells recognize the migration and homing signals,

but lack the specific receptor necessary for correct differentiation and instead differentiate into

melanocytes. In order to further characterize these signals, the effect of various known

differentiation factors could be tested on in vitro cultured precursors for enteric neurons.

6. Clinical implications

A deletion of the long arm of chromosome 10 (10qIl) has been found in a patient with

HSCR (Martucciello et al .• 1992). Recently, linkage has been established between HSCR and

the c-RET proto-oncogene, which maps to this region (Edery et al., submitted). The finding

that transgenic mice carrying a null mutation of the ret-l gene have total intestinal

aganglionosis, further suggests that the RET-J gene might be involved in HSCR (Dr. V.

Pachnis, personal communication). Members of families in which linkage with c-RET was

established, suffered from isolated aganglionosis with no associated anomalies. In our clinical

study (chapter 3.7.), however, we found that HSCR could also be associated with other

major/minor congenital malformations and/or dysmorphisms. HSCR patients were classified

into five groups based on the spectrum of these associated anomalies, both in sporadic and

familial cases of HSCR. This could indicate that multiple pathogenetic mechanisms are

involved in HSCR and stress~s the importance of precise syndrome delineation and recording

of aganglionosis and/or other anomalies in family members, in large series of prospective

case-control studies.

Search for chromosomal abnormalities using high resolution banding techniques, and

linkage analysis analysis in large pedigrees might lead to the identification of other candidate

regions which might harbor genes involved in HSCR. Chromosomal aberrations already found

to be associated with HSCR are trisomy 21 (associated with Down syndrome), a deletion or

trisomy of 22qIl (Passarge, 1967; Spouge and Baird, 1985; Beedgen et al., 1986), and

microdeletions of the long arm of chromosome 13 (in the regionn 13q22-33) (Sparkes et al.,

1984; Kiss and Osztovics, 1989; Lamont et al., 1989; Bottani et al., 1991). Recently, certain

modifyer genes, influencing the phenotypic expression of the disease and its associated

malformations, have been identified for Neurofibromatosis type I (Easton et al., 1993). To

discriminate between the possibilities of multiple genes for HSCR or one primary gene,

phenotypically influenced by modifyer genes, further research is necessary.

147

3.9. References

Allan. l.J .. and Ncwgrecn. D.F. (1980) The origin and differentiation of enteric neurons of the intestine of the fowl embryo. Am. J, Anat. 157:137-154.

Balling, R., Mutter. G .. Gruss. P .• and Kessel. M. (1989) Craruofo.cial o.bnonnalities induced by ectopic expression of the homcobox gene Hox-I.l in tr:lIlsgenic mice. Cell 58:337-347.

Bccdgen. B" Nutzcn:ldcI. w .. Qucrfe1d. U .• and Weiss-Wichert. P. (1986) Partial trisomy 22 and 11 due to:l paternal 11:22 translocation associated with Hirschsprung diseuse, Bur. J. Pediatr. 145:229-232.

Bockman. D.E.. and Kirby. M.L. (1984) Dependence of thymus development on derivatives of the neural crest. Science 223:498-500.

Bottani. A .. Xie, Y .• Binkcrt. F .. and Schinzel. A. (1991) A case of Hirschsprung disease with a chromosome 13 microdclction, dcl(13)(q32.3q33.2): potential mapping of one disease locus. Hum. Genet. 87:74S~750.

Chisaka. 0 .. and Capeechi. M.R. (1991) Regionally restricted developmental defects resulting from targeted disruption of the mouse homeobox gene Hox~1.5. Nature 350:473-479.

Chisaka. 0 .. Musci. T.S .. and Capecchi. M.R. (1992) Developmental defects of the ear, cranial nerves and hindbrain resulting from targeted disruption of the mouse homeobox gene Hox-1.6. Nature 355:516-520.

Ciment. G .. and Weston. J.A. (1983) Enteric neurogenesis by neural crest-derived branchial arch mesenchymal cells. Nature 305:424-427.

Easton, D.F .. Ponder, M.A., Huson. S.M .. and Ponder. B.AJ. (1993) An analysis of variation in expression of neurofibromatosis (NF) type I (NFl): evidence for modifying genes. Am. J. Hum. GeneL 53:305~313.

Epstein. M.L .. Poulsen, KT .. and Thiboldeaux. R. (1991) Formation of ganglia in the gut of the chick embryo. J. Compo Neurol. 307:189-199.

Fantel. A.G., Shepard, T.H .. Newell-Morris, L.N .. and Moffett. B.C. (1977) Tcratogeruc effects of retinoic acid in pigtail monkeys (M::I.I:aca ncmestrina). Teratology 15:65~72.

Hunt. P .. and Krumi::J.uf. R. (1992) Hox codes and positional specification in vertebrate embryonic axes. Ann. Rev. Cell BioI. 8:227-256.

lto, K" and Sieber-Blum. M. (1993) Pluripotent and developmentally restricted neural-crest-derived cells in posterior visceral arches. Dev. BioI. 156:191-200.

Kalcheim, C .. and LeDouarin. N.M. (1986) Requirement of 0. neural tube signal for the differentiation of neural crest eells into dorsal foot ganglia. Dev. BioI. 116:451-466.

Kalcheim, C .. Barde, Y.A .. Thoenen. H .. and LeDouarin. N.M. (1987) In vivo effcet of brain-derived neurotrophic factor on the survival of developing dorsal root ganglion cells. EMBO J. 6:2871-2873,

Kamm. JJ. (1982) Toxicology. carcinogenicity. and teratogenicity of some orally administered retinoids, 1. Am. Aead. Dermatol. 6:652-659.

Kessel. M .. Balling, R .. and Gruss, P. (1990) Variations of cervical vertebrae after expression of a Hox-J,J transgene in mice. Cell 61:301-308.

Kirby, M,L .. Gale. T.F .. and Stewart, D,E, (1983) Neural crest eclls contribute to normal aorticopulmonary septation. Science 220: 1 059-1 061.

Kiss, P .. and Osztovics, M. (1989) Association of 13q deletion and Hirschsprung's disease. J. Med. Genet, 26;793-794.

Kruse. J .. Mailhammer, R .. Wernecke. H .. Faissner, A .. Sommer, I.. Goridis. C .. and Schachner. M. (1984) Neural cell adhesion molecules and myelin-associated glycoprotein share a common carbohydrate moiety recognized by monoclonal antibodies L2 and HNK-l. Nature 311:153-155.

Kuratani, S.c., and Kirby, M.L (1991) Initial migration and distribution of the cardiac neural crest in the avian embryo; An introduction to the concept of the circumpharyngeal crest. Am. J. Anat. 191:215-227.

Lallier, T .. and Bronner-Fraser. M. (1991) Avian neural crest cell attachment to laminin: involvement of divalent cation dependent and independent integrins. Development 113: I 069-1 084.

148

Lallier, T., Leblanc. G .. Artinger. K .. and Bronner-Fraser. M. (1992) Cranial and trunk neural crest cells use different mechanisms for attachment to cxtracellular matrices. Development 116:531-541.

Lammcr. E.J .. Chcn. D.T.. Hoar. R.M .. Agrush. N.D .. Benke. P.J .. Braun. J.T .. Curry. c.)" Fcrnhoff. P.M .. Grix. A.W .. Lott. I.T .. Richard. J.M .• and Sun. S.c. (1985) Retinoic acid embryopathy. N. EngJ. J. Mee!. 313:837-841.

Lamont. M.A.. Fitchett. M .. and Dennis. N.R. (1989) Intcrstitial deletion of dista113q associated with Hirschsprung's disease. J. Moo. Genet. 26:100-1Q4.

LcDouarin. N.M .. and Teillet. M. (1973) The migration of neural crest cells to the wall of the digestive tract in avian embryo. J. EmbryoJ. expo Morpho!. 30:31-48.

LeDouarin. N.M .. and Teillet. M. (1974) Experimental analysis of the migration and differentiation of neuroblasts of the autonomic nervous system and of neuroectodermal mesenchymal derivatives. using a biological cell marking technique. Dev. BioI. 41:162-184.

LeDouarin. N.M. (1982) The neural crest. Cambridge Univ. Press: Cambridge.

Liang. p .. and Pardee. A.B. (1992) Differential display of eukaryotic messengcr RNA by means of the Polymerase Chain Reaction. Science 257:967-971.

Lufkin. T .. Dierieh. A.. LeMeur. M., Mark. M., and Chambon. P. (1991) Disruption of the Hox-1,6 homcobox gene results in defects in a region corrcsponding to its rostral domain of expression. Cell 66: 11 05-1119.

Lumsden. A" Sprawson. N .. and Graham. A. (1991) Scgmental origin and migration of neural crest cells in thc hindbrain fCgion ofthc chick cmbryo. Development 113:1281-1291.

Marshall. R .. Nonchev, S .. Sham, M.R .. Muchamore. I., Lumsden. A" and Krumlauf. R. (1992) Retinoic acid alters hindbrain Hox code and induces transformation of rhombomcrcs 2/3 into a 415 identity. Nature 360:737-741.

Martucciello. G .• Bicocehi. M.P .. Dodero. P .. Lerone. M .. Cirillo. M,$" Puliti. A .• Gimelli. G .. Romco. G" and Jasonni. V. (1992) Total colonic agangJionosis associated with interstitial deletion of thc long arm of chromosome 10. Pedlatr. Surg. Int. 7:308-310.

Miyagawa-Tomita. S., Waldo. K.. Tomita, R .. and Kirby. M.L. (1991) Temporospatial study of the migration and distribution of cardiac neural crest in quail-ChiCk chimeras. Am. J. Anat. 192:79-88.

Newgreen. D.F .• and Erickson. C.A. (1986) The migration of neural crest cclls. Int. Rev. Cyto!. 103:89-145.

Nodcn, D.M. (1983) The role of the neural crest in patterning of avian cranial skeletal. connective and muscle tissues. Dev. Bio!. 96:144-165.

Parikh. D,H., Tam. P.K.M .. Vanvelzen. D" and Edgar. D. (1992) Abnormalities in the distribution of laminin and collagen type IV in Hirschsprung's disease. Gastroenterology 102:l23(}"1241,

Passarge. E. (1967) The genetics of Rirschsprung's disease. Evidence for heterogeneous etiology and a study of sixty~three families, New Eng!. J. Meel. 276;138~143.

Payette. R.F .. Bennet. G.S .. and Gershon. M.D. (1984) NeurofiJament expression in vagal neural crest-derived precursors of enteric neurons. Dev. BioI. 105:273-287.

Payette. RF .. Tennyson. Y.M .. Pomeranz. H.D .. Pham. T.D .. Rothman. T.P .. and Gershon, M.D. (1988) Accumulation of components of basal laminae: association with the failure of neural crest cells to colonize the presumptive aganglionic bowel of lsIls mutant mice. Dev. BioI. 125:341-360.

Pomeranz. H.D .. and Gershon. M.D. (1990) Colonization of the avian hindgut by cells derived from the sacral neural crest. Dev. BioI. 137:378~394.

Pomeranz, H.D .. Rothman. T.P .. and Gershon, M.D. (1991a) Colonization of the post-umbilical bowel by cells derived from sacral neural crest: direct tracing of cell migration using an intcrcalating probe and a replication-deficient retrovirus. Development 111:647-655.

Pomeranz. H.D .. Sherman. D.L .. $malheiscr. N.R .. Tennyson. V.M .. and Gershon. M.D, (1991b) Expression of a neurally related laminin binding protein by neural crcst-derived cells that colonize the gut: relationship to the formation of enteric ganglia. J. Compo Neurol. 313:625-642.

149

Sechrist. J .• Serbcdzija. G.N .• Scherson. T .• Fraser. S.E .. and Bronner-Fraser. M. (1993) Segmental migration of the hindbrain neural crest docs not arise from its segmental generation. Development 118:691~703.

Serbedzija. C.N .• Burgan. S .. Fraser. S.E .. and Bronner-Fraser. M. (1991) Vitlll dye labeling demonstrates a sacral neural crest contribution to the enterie nervous system of chick and mouse embryos. Development 111:857·867.

Shenefclt. R.E. (1972) Morphogenesis of malformations in hamsters caused by retinoic acid: Relntion to dose and stage at treatment. Teratology 5: 1 03~ 118.

Sparkes, RS .. Sparkes. M,C .. Kalina. R.E .. Pagon. R.A .. Salk, DJ .. and Distcche. C.M, (1984) Separ.:l.tion of retinoblastoma and estemse D loci in a patient with sporadic retinoblastoma and dcl(I3) (q14.1q22.3). Hum. Genet. 68:258~259.

Spouge. D., and Baird. P,A, (1985) Hirschsprung's disease in a large birth cohort, Teratology 32:171-177,

Stem. C.D .. and Canning, D,R. (1990) Origin of cclls giving rise to mesoderm and endoderm in chick embryo. Nature 343:273~275.

Teillet. M.A .. Cochard. p" and LcDouarin, N.M, (1978) Relative roles of the mesenchymal tissues and of the complex neural tube-notochord on the expression of adrenergic metabolism in neural crest cells, Zoon 6:115-122,

Thiery. J,P .. Duband, J.L .. and Delouv6e. A. (1982) Pathways and mechanisms of avian trunk neural crest cell migration and localization. Dev. BioL 93:324-343.

Tucker, G.C .. Ciment. G" and Thiery. J,P, (1986) Pathways of avian neural crest cell migration in the developing gut. Dev, BioL 116:439-450.

Vincent. M .. Duband, J .. and Thiery. J. (1983) A cell surface determinant expressed carlyon migrating avian neural crest cells. Dev, Brain Res. 9:235~238.

Webster, W.S .. Johnston. M,e .. Lammer, EJ .. and Sulik. K.K. (1986) Isotretinoin embryopathy and the cranial neural crest: An in vivo and in vitro study. Dev, BioL 6:211~222.

Wolgemuth. OJ" Behringer, RR., Mostoller. M.P .. Brinster, R.L .. and Palmiter. RD. (1989) Transgenic mice overexpressing the mouse homeobox-containing gene Hox-l.4 exhibit abnormal gut development. Nature 337:464-467,

150

SUMMARY

In this thesis, we studied the development of the ENS and the specific role of the hindbrain

neural crest in this process, in avian embryos. The vertebrate neural crest arises on the dorsal

aspect of the neural tube along the entire antero-posterior axis of the embryo. Cells from the

neural crest migrate throughout the body, and give rise to a large variety of derivatives. The

hindbrain or rhombencephalic neural crest gives rise to the intrinsic component of the ENS,

a complex, integrative system of neurons and supportive cells embedded in the wall of the

gut. The ENS regulates bowel motility and in many aspects resembles the central nervous

system. In chapters 1 and 2, we introduce the neural crest and the enteric nervous system. We

summarize the current know ledge on these structures and compare their embryonic

development in various animal species. This showed that although there are some species

specific variations, the basic principles are very similar.

The rhombencephalon consists of eight consecutive rhombomeres. This segmentation

is also found in its neural crest. Cells derived from the fIrst fIve rhombomeres migrate in two

separate streams and populate the fIrst two pharyngeal arches. Segment identity is established

through the expression of a specific combination of lwx genes within each rhombomere and

its neural crest, providing it with a so-called hox code. Neural crest cells retain this code

during migration.

The cellular and molecular mechanisms of the development of the neural crest of

rhombomeres 6-8 are less well known. In chapter 3, we studied whether the neural crest is

regionally specified with regard to ENS formation. We showed that the enteric ganglia of the

hindgut, and perhaps even of the entire gut. derive from the neural crest adjacent to somites

4-5. Ectomesenchymal derivatives. such as cells of the outflow tract of the heart. and the

mesenchymal component of thymus and parathyroids, derive from the neural crest adjacent

to somites 1-3. Cells from rhombomeres 6-8 migrate either dorsolaterally, entering the third.

fourth and sixth pharyngeal arches. or ventrolaterally through the rostral part of the first five

somites. Neural crest cells in the pharyngeal arches mainly give rise to ectodermal derivatives.

Clonal analysis showed that most pluripotent neural crest cells loose the capacity to

differentiate into serotonergic neurons, upon entry into the pharyngeal arches. We surmise that

ENS precursors migrate along a_ventrolateral pathway and do not enter the pharyngeal arches,

in contrast to other rhombencephalic neural crest cells.

Upon entry into the foregut, cells translocate craniocaudally, either as dispersed cells

or in a network. home at the sites of the myenteric and submucous plexuses and differentiate

into enteric neurons. We studied which cells or tissues within the gut wall provide homing

andlor differentiation signals for neural crest cells. We identifIed a group of mesenchymal

cells, carrying two HNK-l immunoreactive cell membrane glycoproteins of 42 and 44 kD.

151

These cells were only found in those parts of the gut which were not yet colonized by neural

crest cells, or in which neural crest cells did not differentiate into neurons. We surmise that

these cells present signals, recognized by neural crest cells, enabling them to home at specific

sites and differentiate into enteric neurons. Differentiation signals in the postumbilical gut.

however, differ from those in the preumbilical gut, in that only vagal neural crest cells are

able to recognize these. Trunk neural crest cells recognize the migration and homing signals,

but lack specific receptors necessary for correct differentiation and instead differentiate into

melanocytes.

Congenital malformations of the enteric nervous system in humans give rise to

significant morbidity and even mortality in early childhood. The most common fonn is

Hirschsprung disease. in which neurons are lacking in the distal colon (aganglionosis). Study

of Hirschsprung disease could provide insight into the development of the enteric nervous

system. A prerequisite for a molecular genetic approach of congenital malfonnations requires

a precise characterization of patient groups. In a retrospective study of patients with

Hirschsprung disease, we found a pattern of associated anomalies. We showed that there is

no real difference between short- and long-segment disease regarding this pattern of

associated anomalies, the sex-ratios and familial occurrence. A number of chromosomal

abnonnalities have been identified in Hirschsprung patients, indicating that there could be

multiple genes involved in causing aganglionosis. Recently, linkage has been established

between Hirschsprung disease and the c-RET proto-oncogene on the long ann of chromosome

10 (1 Oq 11). Future attention will be focused on the role of this gene or genes in its immediate

vicinity in various groups of Hirschsprung patients.

152

SAMENVATTING

In dit proefschrifi hebben wij de ontwikkeling van het zenuwstelsel van de darm en de

specifieke rol van de rhombencepbale neuraallijst bierin bestudeerd in vogel embryonen. De

neuraallijst in vertebraten ontstaat aan de dorsale zijde van neuraalbuis over de gebele antero­

posterior as van het embryo. Cellen afkomstig van deze ne~raallijst migreren door bet

embryo, en geven aanleiding tot een grate verscheidenheid aan derivaten. De neuraallijst

cellen van de achterhersenen of rhombencephalon vormen de intrinsieke component van

darrninnervatie. een complex, geYntegreerd systeem van neuronen en steuncellen in de

dannwand. Het zenuwstelsel van de darm reguleert de darmmotiliteit en lijkt in vele opzichten

op het centrale zenuwstelsel. In de hoofdstukken 1 en 2, worden de neuraallijst en het

zenuwstelsel van de darm geYntroduceerd. We geven een samenvatting van de huidige kennis

over deze structuren en vergelijken hun embryonale ontwikkeling in een aantal diersoorten.

Deze vergelijking geeft aan dat. hoewel er enkele diersoort-specifieke variaties zijn, de basale

principes sterk overeenkomen.

Het rhombencephalon bestaat uit acht opeenvolgende segmenten of rhombomeren.

Deze segmentatie wordt teruggevonden in de bijbehorende neuraallijst. Cellen afkomstig van

de eerst 5 rhombomeren migreren in twee afzonderlijke stromen en bevolken de eerste twee

kieuwbogen. De identiteit van de afzonderlijke rhombomeren en de bijbehorende neuraallijst

wordt bepaald door de expressie van een specifieke combinatie van hox genen, resulterend

in een zogenaamde hox code. Neuraallijst cellen behouden deze code gedurende hun migratie.

De cellulaire en moleculaire mechanismen van de ontwikkeling van de neuraallijst van

de de rhombomeren 6-8 zijn minder bekend. In hoofdstuk 3 wordt aannemelijk gemaakt dat

de neuraallijst van rhombomeer 6-8 regionaal gespecificeerd is met betrekking tot de

ontwikkeling van de danninnervatie. Wij hebben aangetoond dat de dannwand neuronen in

de einddann, en mogelijk zelfs in de gehele dann, atkomstig zijn van de neuraallijst ter

hoogte van de somieten 4 en 5. Ectomesenchymale derivaten, zoals het uitstroom septum van

het hart en de mesenchymale component van de thymus en bijschildklieren, daarentegen, zijn

voornamelijk afkomstig van de neuraallijst ter hoogte van de eerste 3 somieten. Cellen

atkomstig van de rhombomeren 6-8 migreren of weI dorsolateraal, naar de derde, vierde en

zesde kieuwboog, ofwel ventrolateraal door het rostrale deel van de eerst vijf somieten.

Neuraallijst cellen in de kieuwbogen geven nagenoeg uitsluitend aanleiding tot

ectomesenchymale derivaten. Clonale analyse heeft aangetoond dat de meeste pluripotente

neuraallijst cellen het vermogen verliezen om te differentieren in serotonerge neuronen, na

binnenkomst in de kieuwbogen. Wij veronderstellen dat de precursors v~~r de dannwand

neuronen langs ventrolaterale weg migreren en niet via de kieuwbogen, in tegenstelling tot

andere rbombencephale neuraallijst cellen.

153

Na binnenkomst in de voordarm, verplaatsen neuraallijst cellen zich craniocaudaal,

hetzij afzonderlijk, hetzij in een netwerk, en, na het bereiken van hun bestemming (de plaats

van de plexussen van Auerbach en Meissner), differentieren tot darmwand neuronen. Wij

hebben onderzocht welke cellen of weefsels in de darmwand 'homing' en/of differentiatie

signalen leveren voor neuraalIijst cellen. Wij hebben een groep mesenchymale in de

darmwand geYdentificeerd, die twee HNK-l immunoreactive celmembraan glycoproteYnen van

42 en 44 kD tot expressie brengen. Deze cellen kunnen uitsluitend worden aangetoond in

darmdelen die nog niet door neuraallijst cellen zijn gekoloniseerd, of waarin neuraallijst cellen

niet neuronaal zijn gedifferentieerd. Wij veronderstellen dat deze cellen signalen presenteren,

die herkent worden door neuraallijst cellen en hen in staat stellen hun specifieke bestemming

te bereiken en te differentieren tot darmwand neuronen. Differentiatie signalen in de post­

umbilicale darm verschillen van die in de pre-umbilicale darm en alleen vagale neuraallijst

cellen zijn in staat deze te herkennen. Romp neuraallijst cellen herkennen de migratie en

'homing' signalen, maar missen de specifieke receptoren nodig v~~r correcte differentiatie en

in plaats daarvan differentieren zij tot melanocyten.

Aanlegstoornissen van de darminnervatie bij de mens geven aanleiding tot ernstige

morbiditeit en zelfs mortaliteit op de kinderleeftijd. De meest voorkomende vonn is de ziekte

van Hirschsprung, waarbij darmwand neuronen ontbreken in het distale colon (aganglionosis).

Bestudering van de ziekte van Hirschsprung kan inzicht verschaffen in de ontwikkeling van

de darrninnervatie. Een voorwaarde voor een moleculair genetische benadering van

aangeboren afv.rijkingen vraagt een precieze karakterisering van patienten groepen. In een

retrospectieve studie bij patienten met de ziekte van Hirschsprung, vonden wij een patroon

in de geassocieerde afwijkingen. We hebben aangetoond dat er geen wezenlijk verschil is

tussen kort- en lang-segment aganglionosis wat betreft dit patroon van geassocieerde

afwijkingen, de geslachtsverhoudingen of het familiaal voorkomen. Een aantal chromosomale

afwijkingen zijn geYdentificeerd in patienten met de ziekte van Hirschsprung. Dit suggereert

dat er meerdere genen betrokken zouden kunnen zijn bij het ontstaan van aganglionosis.

Recent is er koppeling aangetoond tussen farniliale gevallen van de ziekte van Hirschsprung

en het c-RET proto-oncogen gelegen op de lange arm van chromosoom 10 (10qll). Studie

naar de reI van dit gen of nabij gelegen genen in verschillende groepen patienten met de

ziekte van Hirschsprung en in proefdieren vonnt een aandachtsveld voor toekomstig

onderzoek.

154

Curriculum Vitae

PersonaIia:

Naam:

Geboren:

Gehuwd:

Maria Josepha Hubertina van der Sanden

24 oktober 1962, te Geldrop

Wilhelmus Mathias Jacqelina Peters, 12 juli 1984

Opleiding en diploma'S:

1981 Eindexamen Gymnasium S, Augustinianum, Eindhoven

1984 Kandidaatsexamen Medische Biologie (B5'), Rijksuniversiteit Utrecht

1986 Proefdierkundige (ex. art. 9 Wet op de dierproeven)

1987

1987

1988

1988

Posities:

1987-1988

1988-1993

Stralingsdeskundige (4B)

Doctoraalexamen Medische Biologie

Hoofdvakken:

Identificatie van de verschillende subpopulaties cellen in het

mammacarcinoom van de kat zowel in vivo als in vitro, m.b.v.

polyc1onale en monoclonale antilichamen. bij de vakgroep Veterinaire Pathologie, begeleiding Prof. Dr. E. Gruys.

Functionele studie naar de toxiciteit van deoxyadenosine voor

adenosine-deaminase defiente B lymfocyten. tijdens activatie en

in rust, bij de vakgroep Immunologie. Wilhelmina Kinderziekenhuis Utrecht. begeleiding Dr. B.M. Zeegers.

Bijvak.: Algemene Dierkunde. projectgroep experimentele

embryo1ogie - Prof. Dr. W.L.M. Geilenkirchen

Cambridge Proficiency Certificate, grade A

AIO-examen statistiek

Tijdelijke aanstelling in het kader van een pilot-studie naar het effect

van stress op het immuunsysteem, bij de vakgroep Immunologie van

het Academisch Ziekenhuis Utrecht, Prof. Dr. R. Ballieux.

Thesis project getiteld: 'The hindbrain neural crest and the development

of the enteric nervous system: Dit onderzoek maakt deel uit van het

Medisch Genetisch Centrum Zuid West Nederland (sectie 2D) en is

bewerkt binnen een samenwerkingsverband van het Instituut

155

1994-1996

156

Kinderheelkunde en het Instituut Celbiologie en Genetica, Erasmus

Universiteit Rotterdam, onder de verantwoording van Prof. Dr. le. Molenaar en Dr. C. Meijers.

Post-doc project getiteld:'The role of the ret-l gene in the formation

and malformation of the enteric nervous system' (SSWO-project). Dit

onderzoek wordt bewerkt binnen hetzelfde samenwerkingsverband als

het thesis project.

List of publications

Peters-van der Sanden, MJ,R, Luider, T.M .• van der Kamp, AW.M., Tibboel, D., and

Meijers, C. (1993) Regional differences between various axial segments of the avian neural

crest regarding the formation of enteric ganglia. Differentiation 53:17-24.

Peters-van der Sanden, M.J.H., Kirby, M.L.. Gittenberger-de Groot, AC., Tibboel, D., Mulder,

M.P., and Meijers, C. (1993) Ablation of various regions within the avian neural crest has

differential effects on ganglion formation in the fore-, mid-. and hindgut. Dev. Dyn., 196: 183-

194.

Luider, T.M., Peters-van der Sanden, M.J.H., Molenaar, J.e., Tibboel, D., van der Kamp.

A.W.M., and Meijers, e. (1992) Characterization of HNK-l antigens during the formation of

the avian enteric nervous system. Development 115:561-572.

Meijers, e., Peters-van der Sanden. MJ.H., Tibboel, D., van der Kamp, A.W.M., Luider.

T.M .• and Molenaar, J.C. (1992) Colonization characteristics of enteric neural crest cells:

embryological aspects of Hirschsprung's disease. J. Pediatr. Surg. 27:811-814.

Nishijima. E., Meijers, C., Tibboel. D., Luider, T.M., Peters-van der Sanden, MJ.R. van der

Kamp, AW.M., and Molenaar, J.e. (1990) Formation and malformation of the enteric nervous

system. J. Pediatr. Surg. 25:627-631.

157

Dankwoord

Tot slot wi! ik graag iedereen bedanken die op enigerlei wijze een bijdrage hebben geleverd

aan de tot stand korning van elit proefschrift. Want, ondanks het feit dat er slechts een naam

prijkt op het titelblad, berust het verschijnen van een proefschrift op teamwork. Een aantal

mensen wil ik hier met name noemen.

AlIereerst wi! ik Prof. Molenaar bedanken voor de roij geboden gelegenhcid om rut onderzoek te kunnen verrichten en het in rnij gestelde vertrouwen dat het mede mogelijk heeft

gemaakt dit onder.lOek binnen zijn afdeling verder voort te zetten. Een bijzonder woord van dank wil ik richten tot Dr. C. Meijers. Carel, jouw hulp bij

het tot standkoming van dit proefschrift is niet in cen paar woorden te beschrijven. Jouw

ideeen en enthousiasme zijn vaak. cen stimulans veor rni j geweest om door te gaan, voora!

ook in de peri ode dat het onderzoek niet zo wilde lukken. Plezierige herinneringen heb ik aan de (zeer) langdurige werkbespreking in Heidelberg. Ik waardeer het dan ook zeer dat jij co­promoter wilt zijn bij mijn promotie, en ik: hoop dat onze verdere samenwerking even

plezierig en succesvol zal zijn.

I would like to thank Prof. Dr. Med. B. Christ for his agreement to act as external

examiner. Ook wil ik Prof. A. Gittenberger-de Groot en Prof. Dr. F. Grosveld bedanken voor

hun bereidheid zitting te nemen in de leescommissie.

Dr. M. Wilke en Dr. M-J. Vaessen wi! ik bedanken omdat zij rnij als paranirufen

terzijde willen staan in deze voor mij zo enerverende periode. Martina, Marie-losee, jullie

vriendschap, mede door gedeelde interesses buiten de werksfeer, betekent veel voor mij.

Martina, ik hoop dat wij nog lang samen al hupsend aan onze conditie zullen werken. Marie­

los6e, ondanks dat onze koffie-gesprekken Diet altijd door iedereen werden gewaardeerd, hoop

ik toch dat wij deze nog vaak zullen voeren.

Arthur van der kamp, Theo Luider en Dick Tibboel dank ik: voor hun begeleiding in

de begin periode van mijn onderzoek, voora! ook in de tijd dat Carel in Heidelberg was.

Mijn lab- en kamergenoten wi! ik: bedanken voor de plezierige samenwerking en de

gezellige werksfeer. Allereerst de mensen van mijn huidige lab: Sylvia S., Laurens. Marie­

Josee, An, Nynke, Sylvia D., Dirk-Jan, Clara, Sjaak en Gaetano. Maar ook rnijn ex-Iabgenoten

van 710, Rien, Riny en Ton, en het EM lab, Han, Pim V., Rob en Lies-Anne, en de bewoners

van de CF-kamer, Bob, Jan, Hikke en Pim F ..

I would like to thank Prof. M.L. Kirby for her pleasant and fruitful collaboration

which resulted in the work described in chapter 3.2.

Ilse van Haperen-Heuts wil ik: bedanken voor de inwijding in de mysteries van het

kippe-embryo (het vrijprepareren van 12 buizen per dag was in het begein een hele klus).

Verder wi! ik: de studenten en stagiaires - Nathalie Bravenboer, Irma van de Velde, Ragonda

Souren, Willeke Verheul, Iris Nurmohamed. Leon Verhoog en Anja Raams - bedanken voor

159

hun bijdrage aan mijn onderzoek.

Monica Menting dank ik v~~r de advanced course in de constructie van kip-kwartel

chimeren. Samen waren we in staat te frustraties van te jonge of te Dude embryos en

onbevruchte eieren te doorstaan. De overige medewerkers van het lab anatomie en

embryologie in Leiden wi! ik bedanken v~~r hun gastvrijheid en menige prettige koffie en

lunch gesprekken.

Verder wi! ik alle nog Diet genoemde medewerkers van de afdelingen Celbiologie en

(Klinische) Genetica bedanken v~~r de plezierige samenwerking. Een aantal mensen die

onmisbaar zijn voor het functioneren van een afdeling wil ik hier met name noemen: Piet

voor de technische assistentie; Jan Jos v~~r de computerondersteuning; Tom, Mirko en Ruud

voor de vele (?) foto's en dia's; Marike, Jeanette en Rita voor de secretarie!e hulp; Jopie, EUy

en lake voor het altijd weer schone glaswerk, en Rein, Melle en Mieke voor de zorg voor alle

bestellingen.

Mijn ouders wi! ik bedanken voor hun nimmer aflatende steun en belangstelling. Tot

slot maar zeker niet in de laatste plaats, Wil. Het belang van een thuishaven waar je je kunt

ontspannen en het werk af en toe kunt relativeren valt Diet te onderschatten.

160